ICCCI Conference Proceedings

13 Edition
th
IEEE S ponsored
2023 International Conference

on Computer Communication
and Informatics
Date: 23rd - 25 th, January 2023
Venue: SSIET College Campus
Machine Learning &

Artificial Intelligence Robotics
Embedded Systems Industrial

& IOT Automation
Bitcoin,
block chain & Ethical Hacking
Cryptocurrency and Network Security
Technical Sponsors:
Approved by AICTE, New Delhi Affiliated to Anna University, Chennai
An Autonomous Institution
L&T Bypass, Chinniampalayam, Coimbatore - 641 062. Tel : +91-422-2369900 to 2369999 (100 Lines)
Web : www.siet.ac.in
ISBN: 979-8-3503-4821-7
2023 International Conference on Computer Communication and Informatics
January 23-25, 2023, Coimbatore, India
TECHNICAL SPONSORS
Copyright © 2023 by the Institute of Electrical and Electronics Engineers, Inc. All rights
reserved.
Copyright and Reprint Permission
Copyright and Reprint Permission: Abstracting is permitted with credit to the source. Libraries
are permitted to photocopy beyond the limit of U.S. copyright law for private use of patrons
those articles in this volume that carry a code at the bottom of the first page, provided the per-
copy fee indicated in the code is paid through Copyright Clearance Center, 222 Rosewood
Drive, Danvers, MA 01923. For reprint or republication permission, email to IEEE Copyrights
Manager at [email protected] rights reserved. Copyright ©2022 by IEEE.
i
ISBN: 979-8-3503-4821-7
Sri Shakthi Institute of Engineering and Technology (SIET) was established in the year 2006
by Dr.S. Thangavelu, with the zeal to provide quality Engineering Education to the young minds and
to make them innovative employable Engineers.
Sri Shakthi is accredited by NAAC with A grade and also SIET is the youngest institution in
India, accredited by National Board of Accreditation (NBA) for four courses namely Agri, Mech,
ECE, EEE, CSE and IT. We have been consecutively recognized as the best industry connect institute
with platinum ranking by the AICTE-CII survey of Industry-linked technical Institutions. SriShakthi
symbolizes 'creative & progressive power' of dynamic youth and is ranked among top 10 percent of
3000 colleges across India to receive National Employability Award and The Times Group Award for
Excellence in Education. We are the pioneer Institute in India, chosen by Indian Electronics &
Semiconductor Industries Association to launch their premier courses on VLSI design and embedded
product design
The inspirational leaders such as Padmashri A. Sivathanu Pillai, Padmashri R.M. Vasagam,
MylswamyAnnadurai, Dr. Sandeep Garg, P. VenkatRangan, Mr. Sanjeev Keskar, Mr.
SrikantanMoorthy, A.K.Pattabiraman, Mr. Varadharajan, Ms. Hema Gopal, Mr. K. Ganesan,
Madhusudan Atre, Mr. VivekPawar, Kamesh Namuduri, Mr. Jayaram Pillai, Mr. Veerappan, Mr.
Parthasarathy, Mr. VA Shiva Ayyadurai, Kiran Bedi, Prof. John A Davis, KaviPerarasuVairamuthu,
Mr. Chetan Bhagat, Mr. Rajesh Nair, Mr. Kannan Ramamani and Mrs. Anuradha Srinivasan have
visited our campus to inspire our students
The institute is also collaborated with worldwide universities and industries to support our
mission. Oracle, CISCO, National Instruments, Cadence, Xilinx, Infosys, Wipro, MindTree, AMI,
Siemens, Dassault Systems, TTK prestige, HP Enterprise, Virtusa Polaris, Gyan Matrix, IESA,
NASSCOM, IEI, ISTE, IEEE and ITB are few among them.
The institute currently offers thirteen bachelors degree programs in the field of Agricultural,
Bio Medical, Bio Technology, Electrical and Electronics, Electronics and Communication, Computer
Science, Cyber Security, Artificial Intelligence and Data Science, Artificial Intelligence and Machine
Learning, Food Technology, Information Technology, Mechanical, and Civil and master degree
programs in the fields of VLSI Design, Engineering Design, Structural Engineering, Computer Science
and Engineering, Embedded System Technologies.
ii
ISBN: 979-8-3503-4821-7
Message from Chairman

The ICCCI 2023 is the 13th edition, conducted every year in the month of January by Sri
Shakthi Institute of Engineering and Technology and the forum of great novelty with the
views of initiatives. More than 700 research papers are received every year and around 350
are selected for presentation after the review by the experts of research. This year 475 papers
has been registered for presentation and the research articles are received from Australia,
Bangladesh, China, Taiwan, Malaysia, United Kingdom and also from various reputed
Institutions from all over India
Research papers are received in the recent technologies like Machine learning, Artificial
Intelligence, Embedded systems, IoT, robotics, Industrial Automation, Bitcoin, Blockchain,
Crypto currency, Ethical hacking, and Network Security.
The conference of this kind would provide a great platform for the higher educational
institutions to understand the various real time industry-based research needs. In turn, this
would provide a road map to enhance the overall quality of higher education.
Another important aspect is the interdepartmental research. In today’s world of internet of

things, augmented reality, virtual reality and machine learning, the interdepartmental
research has become imperative and felt necessary.
Further, research should be focused to improve the agricultural production and the overall
quality of life by avoiding all types of pollutions. The computer, communication and
informatics would play a major role to full fill the above dream.
The main focus of conference is aimed to take up research in the growing, need based, market
segments, where there is a lot of scope for bringing out high-tech products and for registering
patents. The papers presented in the conference would be sent to the IEEE for indexing in
the IEEE Xplore. Apart from this the selected papers would be sent to Web of Science and
Scopus Indexed Journals.
My best wishes to the conference.

Dr.S. Thangavelu,
Chairman,
Sri Shakthi Institute of Engineering and Technology.
iii
ISBN: 979-8-3503-4821-7
Message from Principal
I am delighted to meet you through this page. Education is not only an act of acquiring
knowledge but learning a skill to lead life and grooming ones personality. Education of the
highest order aims at guiding, inspiring, motivating and leading young men and women to
become successful leaders to serve the country better. Research is the key parameter to
promote the individuality to horizon. In order to create the best Engineers, the college has
been providing environment to enhance the research activities even from the budding student
Engineers since its origin.
2023 International Conference on Computer Communication and Informatics (ICCCI), an
advent venue of the Engineering researchers is planned this year with all the victorious
establishments of the last year. I hope this forum will provide enlightening contributions from
researchers of Engineering filed this year too.
With immense pleasure, I welcome all the contributors for ICCCI 2023.
Dr.A.R.Ravikumar
Principal,
Sri Shakthi Institute of Engineering and Technology.
iv
ISBN: 979-8-3503-4821-7
Preface
The 2023 International Conference on Computer Communication and Informatics
(ICCCI 2023) aims to provide an outstanding opportunity for both academic and industrial
communities alike to address new trends, challenges and emerging technologies on topics
relevant to today’s technological advancements in the areas of Computer, Communication
and Informatics. The conference features invited talks and highly commended paper
presentations. The vision of ICCCI 2023 is to foster communication among researchers and
practitioners working in a wide variety of areas in communication and informatics with a
common interest.
On this occasion, I feel immensely happy and proud in welcoming everyone for the
conference and I would like to thank everyone who has shown interest in participating the
conference, the internal and international technical committee members for their valuable
time and all the faculty members who supported. The success of the conference depends
ultimately on many teachers who have worked with us in planning and organizing both the
technical program and supporting social arrangements. Generous support for the conference
is provided by IEEE. Finally, I would like to thank Chairman, Secretary, Joint Secretary,
Principal, and Dean Academics of Sri Shakthi Institute of Engineering and Technology for
their strong support, motivation and valuable guidance for the successful conduct of this
grand event.
Dr.A.Kalaiarasi
Conference Chair, ICCCI 2023
v
ISBN: 979-8-3503-4821-7
Scope of ICCCI 2023
India has the greatest potential to address the global challenges in the engineering and design
arena. 400+ global Engineering R&D companies already have their presence in our Country.
The major part of young talents with diverse skill set, along with growing academic and
research infrastructure and a progressive policy environment creates more research activities
in the country. The new dimension of globalization of R&D and FDI support spurs the
momentum of growth opportunities for the R&D activities in Engineering.
The vision of ICCCI 2023 is to develop communication among researchers and practitioners
working in a wide variety of areas in communication and informatics with a common
interest.
The 2023 International Conference on Computer Communication and Informatics (ICCCI

2023) aims to create a forum for the people involved in Engineering R&D activities to
discuss about:
• Unique Opportunity to Partner for Maximum Visibility & Community Engagement

• An outstanding opportunity for both Academic and Industrial Communities to
debate and address new trends and challenges with
• Focused sessions for specific verticals
• Master Classes for hand-holding on issues on the ground
• Share common challenges
• Avenue to ideate about emerging Technologies on topics relevant to today’s fast
moving areas of Computer, Communication and Informatics by Explore
strategies
• Listen to the best of engineering minds
• Create a common framework for the future
• Forum to collaborate with
• Research scholars and PG students
• Business Connect
• Global leaders
• Academia
• Analysts
• Peers
vi
ISBN: 979-8-3503-4821-7
Sri Shakthi Institute of Engineering & Technology, Coimbatore 641062 | 23 - 25, January 2023
Date Slot Time Speakers
Conference Inaugural Session
Xavier Fernando, PhD, P. Eng.,

Professor, Toronto Metropolitan University
Director, Ryerson Communications Lab
Associate Editor, IEEE IOT Journal
IEEE Canada Region 7 Vitality Coordinator
23-Jan-2022 FN 09 .00 am – 10.00 am Dr Nalinda Somasiri, BEng (Hons), MSc, PhD

Head of Programme - Computer Science
London Campus
York St John University
Dr.K.Porkumaran
Chairman-IEEE Madras Section
Primcipal and Professor
Sri Sai Ram Engineering College, Chennai
Dr.Ramesh Kumar Ayyasamy

Professor, Faculty of Information and
Communication Technology,
Universiti Tunku Adbdul Rahman, Malaysia
Dr.Fursan Thabit
Professor, Cloud Computing and IoT
Faculty of Engineering,
Ege University, Turkey
Conference Valedictory Session
Dr.Renuga Karthika
Professor & Dean of Student Affairs
25-Jan-2022 AN 04.30 pm- 05.00 pm Department of Information Technology
PSG College of Technology
Coimbatore
vii
ISBN: 979-8-3503-4821-7
Committees
Chief Patron
Dr.S.Thangavelu,
Chairman,
Sri Shakthi Instittute of Engineering and Technology
Patrons
Shri. T. Dheepan,
Vice Chairman,
Shri. T. Sheelan,
Joint Secretary,
Dr.A.R.Ravikumar,
Principal
Conference Chair
Dr.A.Kalaiarasi,
Associate Professor,
Department of EEE,
Publication Chair
Dr.K.E.Kannammal,
Head of the Department
Computer Science Engineering
Technical Program Chair

Dr.S.Bhavani,
Eletronics and Communication Engineering
Conference Treasurer
Dr.S. Prakash
Information Technology
Sri Shakthi Institute of Engineering and Technology
viii
ISBN: 979-8-3503-4821-7
INTERNATIONAL ADVISORY COMMITTEE
Prof. Dr. Manuel Cedillo-Hernandez, Universidad Nacional Autonoma de Mexico

Dr. Masahiro Fujita, Professor, University of Tokyo, Japan
Dr.Vijayaragavan Viswanathan, Researcher, CERN - Jablotron Alarms, CZECH REPUBLIC
Prof. Shaneel Narayan, Unitec Institute of Technology, Auckland
Professor Vahab Pournaghshband, California State University, Northridge, United States
NATIONAL ADVISORY COMMITTEE
Dr. K. Porkumaran, Chairman, IEEE Madras Section

Dr .P. Sakthivel, Vice Chairman Industry, IEEE Madras Section
Dr. R. Hariprakash, Secretary, IEEE Madras Section
Dr. S. Radha, Treasurer, IEEE Madras Section
Dr. N. Kumarappan, Immediate Past Chairman, IEEE Madras Section
Dr. S. Brindha, Member, IEEE Madras Section
Dr. D. Nirmal, Member, IEEE Madras Section
Mr. M. Arun, Member, IEEE Madras Section
Dr. R. G. Sangeetha, Member, IEEE Madras Section
Mr. B. Ashvath, Member, IEEE Madras Section
Dr. V. Nagarajan, Member, IEEE Madras Section
Dr. S. Rathinavel, Member, IEEE Madras Section
Dr. T. Thyagarajan, MIT, Chennai
Mr. Ram Satagopan, Director- Control Engg., Lennox International
Mr. Nagaraj Subramanyam, Director, ekLakshya VLSI R & D Centre
Dr. K. Giridhar, Indian Institute of Technology Madras, Chennai.
Mr. Kumar Madhavan, Enterprises Architect,
Mr. Ashwin Sadasivam Sundar, Director global head core practice, Virtusa
ix
ISBN: 979-8-3503-4821-7
INTENATIONAL REVIWERS
S.No NAME DESIGNATION COLLEGE NAME

National Autonomous University of
1 Manuel Cedillo-Hernandez Assistant Professor
Mexico
2 M.Sadish Sendil Assistant Professor Sur University College
3 J.Bhuvana Assistant Professor King Khalid University
4 Ramu Avanashiappan Assistant Professor

Adigrat University
5 R Anandhakumar Professor
Adigrat University
6 Chandrasekharan Nataraj Senior Lecturer

Asia Pacific University
7 Bela Genge Assistant Lecturer Petru Maior University of Tg. Mures
8 Xiao-Zhi Gao Professor

Xiao-Zhi Gao
9 Seyed M. Buhari Associate Professor

King Abdul Aziz University
10 Rosula Reyes Associate Professor

Ateneo de Manila University
11 P.D.D.Dominic Associate professor

Universiti – Teknologi
12 Wael Mohamed Khedr Assistant Professor

Majmaah University
13 Ahmed Hassan Mohammed Hassan Assitance Professor

University of Blue Nile
14 Hamid Ali Abed AL-Asadi Professor and Head Basra University
15 Yong Li Associate Professor

Tsinghua University
Khandoker Mohammed Mominul Shahjalal University of Science and
16
Haque Principal Engineer Technology
17 Kalai Anand Ratnam Senior Lecturer
18 Selvakumar Samuel Senior Lecturer

University of Babylon Republic of
19 Eng.Ahmed Kadhim Hussein
Assistant Professor Iraq,
20 S. Dheva Rajan Assistant Professor
Al Musanna College of Technology
Nanhua University, College of Science
21 Jue-Sam Chou
Professor and Technology
22 A. Ramu Assistant Professor
Adigrat University
23 Gwanggil Jeon Assistant Professor

Incheon university
24 Anand paul Associate Professor

Kyungpook national University
25 Karen Cheung Professor The University of British Columbia
x
ISBN: 979-8-3503-4821-7
26 Richi Nayak Associate Professor

Queensland University of Technology
27 Nakhoon back Professor

Kyungpook national University
28 Tor Aamodt Professor The University of British Columbia
29 Sigurd Meldal Professor

San Jose State university
30 Tilahun Mamuye Gidey Technical Assistant

Mekelle University
31 Xavier Fernando Professor

Ryerson University
32 Alaa Hussein Al Hamam Assistant Professor

Amman Arab University
33 Hamid R Arabia Professor

University of Geargia
34 Ahmed F.Zobaa Senior Lecturer

brunel university
35 Dr. Arthur James Swart Associate Professor Central University of Technology
36 Dr. Santolo Meo Professor FEDERICO II University
37 Dr. Ismail Musirin Professor Universiti Teknologi Mara
38 Dr. Ing. Erick Vincent Mgaya Senior Lecturer Arusha Technical College
Al Ain University of Science and
39 Dr. Rafa E.Al.Qutaish Deputy Dean
Technology
40 Dr. Moses Garuba Professor Howard University
41 Yuen , Henry Assistant Professor columbia university
42 Sanjeev Arora Professor Princeton University
43 Stein, Clifford Professor Columbia University
44 Ryan Adams Professor Princeton University
45 Sathish Gopalakrishanan Associate Professor The University of British Columbia
46 David August Professor Princeton University
47 Daniel Alexander Professor University College London
48 Mohan Kankanhalli Professor National University of singapore
49 David Barber Professor University College London
50 Alessandro Abate Professor University College London
51 Stephane Bressan Associate Professor National University of singapore
52 Zemel, Richard Professor columbia university
xi
ISBN: 979-8-3503-4821-7
53 Harold SOH Assistant Professor National University of singapore
54 Tomaso Aste Professor University College London
55 Austeng, Andreas Professor University of Oslo
56 Bratteteig, Tone Professor University of Oslo
Norwegian University of Science and

57 Pål Risan Assistant Professor
Technology

58 Suzan Bulduk Professor
Technology
59 Boualem Benatallah Scientia Professor University of New South Wales
60 Vijay Bhargava Professor The University of British Columbia
61 Mahbub Hassan Professor University of New South Wales
62 Elaine Stratford Professor University of Tasmania
63 Andrew McMinn Professor University of Tasmania
64 Dr Shady H E Abdel Aleem Associate Professor Higher Institute of Engineering
65 Dr.Dhafer Almakhles Associate Professor Prince Sultan University
66 Dr.Jagabar Sathik M Associate Professor Prince Sultan University
67 Dr.Imran Baig Professor Dhofar University
68 Dr.Prajoona Valsalan Professor Dhofar University
69 Dr Ahmad I Omar Professor The Higher Institute of Engineering
70 Jesus Calvino Fraga Professor The University of British Columbia
71 Wong Wai kit Professor Multimedia University
72 Dr.Ayoobkhan Professor South Eastern University
73 Dr Thayabaran Kathiresan Professor University of Zurich
74 Dr Ali Ameen Professor Lincoln University
75 Dr.Barath Narayanan Adjunct Professor Dayton University
76 Dr Selvakumar manickam Associate Professor Universiti Sains Malaysia
77 Sakkaravarthi Ramanathan Professor Vanier College
78 Juri Jatskevich Professor The University of British Columbia
xii
ISBN: 979-8-3503-4821-7
79 Carol Jaeger Professor The University of British Columbia
80 Albregtsen, Fritz Professor Emeritus University of oslo
81 Fawaz S Al Anzi Professor Rensselaer Polytechnic Institute
82 Lau Bee Theng Senior Lecturer Swinburne University of Technology
Chittagong university of Engineering and

83 Kaushik Deb Professor
Technology
84 Alexandra Fedorova Professor The University of British Columbia
85 Kevin Elphinstone Assistant Professor University of New South Wales
86 Y.Christine Chen Associate Professor The University of British Columbia
87 Iskandar Shah Salleh Assistant Director

Nanyang Technological University
88 Kjell Sand Professor
Technology
89 Dr.B.Vivekanandam Deputy Dean Lincoln University
Berlin School of Business and Innovation
90 Dr. Duraisamy Balaganesh Lecturer
GmbH
91 Dr. Hishamuddin Bin M.Salleh Lecturer Lincoln University
92 Dr.Midhunchakkaravarthy Dean Lincoln University
93 Dr.Divya Midhun Chakkaravarthy Director Lincoln University
94 Kambombo Mtonga Lecturer University of Rwanda
95 Tiwonge Manda Senior Lecturer University of Malawi
96 Steven Wandale Lecturer University of Malawi
97 Innocent Kadaleka Phiri Lecturer University of Malawi
98 Thokozani Felix Vallent Lecturer University of Rwanda
99 Willie Kasakula Developer National Integrated Technologies Limited
100 Sarah Nawoya Lecturer University of Rwanda
101 MUTONI brice quercy Lecturer University of Rwanda
102 asya mgeni Lecturer University of Rwanda
103 angel gabriel Lecturer University of Rwanda
104 Halima Mohamed Lecturer University of Rwanda
105 Oluwatobi Oyinlola Lecturer University of Rwanda

xiii
ISBN: 979-8-3503-4821-7
106 Paul Stone Macheso Lecturer University of Malawi
107 Yamikani Phiri Lecturer University of Rwanda
108 Lawrence Byson Lecturer University of Rwanda
109 Yu Chu Tian Professor Queensland University of Technology
110 Alexander Associate Professor University of North California
111 Awais Ahmad Associate Professor University of Milan
112 Dr. Adem Tekerek Associate Professor Gazi University

University of Doha for Science &
113 Dr. Mueen Uddin Associate Professor
Technology
114 Dr. Dinesh Mavaluru Assistant Professor Saudi Electronic University
115 Dr. Xiaohui Yuan Associate Professor University of North Texas
116 Dr. Divya Prakash Shrivastava Associate Professor Higher College of Technology University
117 Dr. Mariofanna Milanova Professor University of Arkansas
Thomas J Watson College of Engineering

118 Mohammad T. Khasawneh Professor
and Applied Science,
119 Dr. Ramon Barber Castano Director Universidad Carlos III de Madrid
120 Dr. Sheng-Lung Peng Professor National Taipei University of Business

Ladoke Akintola University of
121 Stephen Olatunde Olabiyisi Professor
Technology
122 Dr. Xavier Fernando Professor Ryerson University
xiv
ISBN: 979-8-3503-4821-7
IEEE REVIEWERS

Aarupadai Veedu Institute Of
1 Dr.D.Vijendra Babu Professor
Technology
Vel Tech Institute Of Science And
2 Dr.N.Vinodhkumar Associate Professor
Technology
Mallareddy Institute Of Engineering
3 Dr. S. Prabaharan Associate Professor
And Technology
4 Dr. G. Babu Professor Easwari Engineering College
5 Anbumani K Associate Professor Sri Sairam Engineering College
6 Dr.N.Shanmugasundaram Professor Sri Eshwar College Of Engineering
7 Dr.Ramsamy S Assistant Professor Ksr College Of Engineering
8 N.Malmurugan Professor Mahendra Engg College
9 D Nirmal Professor Karunya University
Sri Ramakrishan Engineering
10 K.Balamurugan Assistant Professor
College
11 Reddy G Senior Professor Vit University
12 Srinivasan Selvan Associate Professor Alpha College Of Engineering
13 C.Sathiya Kumar Associate Professor Vit University
14 Madheswaran M Professor Mahendra Engineering College
K.S.Rangasamy College Of
15 Velumani Assistant Professor
Technology
16 S.Elangovan Professor Janson Institute Of Technology
Kumaraguru College Of
17 Marimuthu Ramalatha Professor
Technology
18 Michael T Associate Professor Karunya University
19 Suresh Sankaranarayanan Professor Srm University
St.Xavier's Catholic College Of
20 Raju Darwin Jose Associate Professor
Engg
21 Pai G A Vijayalakshmi Professor Psg College Of Technology
22 Srikrishna Subramanian Professor Annamalai University
23 K .Porkumaran Director Sairam Group Of Institutions
24 Bhaskar Vidhyacharan Professor Srm University
25 P Swarnalatha Associate Professor Vit University
26 Tripathy.B K Professor Vit University
27 Venkataraman Saravanan Professor Dr. Ngp Institute Of Technology
28 S.P.Vimala Associate Professor
College
29 Raghavan Dhanasekaran Professor Syed Ammal Engineering College
30 M.Senthamil Selvi Professor
College
31 P.Sakthivel Associate Professor Anna University
32 P.Ganeshkumar Assistant Professor Anna University Regional Campus
33 Dr S Amutha Associate Professor Saveetha Engineering College
34 Niha K Assistant Professor Saveetha Engineering College
xv
ISBN: 979-8-3503-4821-7
35 Dr. V. Karthikeyan Professor Amet Deemed To Be University,

36 Dr. P. S. Mayurappriyan Professor
Technology
Adama Science And Technology
37 Balachandran Ruthramurthy Professor
University
Karunya Institute Of Technology &
38 D.S.Shylu Sam Associate Professor
Sciences
B.S.Abdur Rahman Crescent
39 P.K.Jawahar Professor
Institute Of Science & Technology
40 Subbanagounder Arumugam Professor Nandha Engineering College
41 Dr.S. Elangovan Professor Jansons Institute Of Technology
42 M.Arun Associate Professor Panimalar Institute Of Technology
Oil And Natural Gas Corporation
43 Nvln Prasad General Manager
Limited
44 Dr Ilangokaruppasamy Assistant Professor(Sl) Amrita Vishwa Vidyapeetham
Sri Ramakrishna Engineering
45 Dr.K.Srinivasan Associate Professor
College
46 Dr Senthamil Selvi Professor
College
47 Dr.S..Vimal Professor
College
48 Dr T Thiagarajan Professor Madras Institute Of Technology
49 Dr S Ranganatha Muthu Professor Ssn College Of Engineering
50 Dr Pa Manoharan Scientist Drdo
51 Dr.B.Paramasivam Professor National Engineering College
52 T S Rangarajan Senior Vice President Tata Consultancy Services
53 Dr S Sundaresh Distinguished Fellow Drdo
54 Dr R Dhanasekaran Director Research Syed Ammal Engineering College
55 Dr N.R Alamelu Professor
College
56 Dr R Gunabalan Associate Professor Vit University
57 H R Mohan Engineer Avp Systems
58 Dr S Malathi Professor Panimalar Engineering College
59 Dr.L.Raja Associate Professor Sri Eswar College Of Engineering
60 G.Manimala Associate Professor Sri Sai Ram Engineering College
61 V Lalitha Associate Professor Sri Sai Ram Engineering College
62 V Kavitha Associate Professor Sri Sai Ram Engineering College
63 D Sathishkumar Associate Professor Sri Sai Ram Engineering College
64 N.Gopinath Associate Professor Sri Sai Ram Engineering College
65 Gajula Ramana Murthy Professor Alliance University
K.S.Rangasamy College Of
66 Professor
Dr.R.Gopalakrishnan Technology
Muthayammal Engineering College
67 Professor
Dr.M.Madheswaran (Autonomous)
68 Dr. Pandian Gurusamy Senior Manager (Retd) Iocl Vishnu Institute Of Technology
xvi
ISBN: 979-8-3503-4821-7
Navaneethakrishnan Kumaraguru College Of

69 Professor
Ramanathan Technology
Amrita Vishwa Vidyapeetham
70 Professor
Dr. Prashant R. Nair Coimbatore
71 Professor
Bharathipriya C Technology
Kongu Engineering College
72 Professor
Nithyavathy N Perundurai Tamilnadu India
73 Professor
Dr.D.Vijendra Babu Technology
74 Dr N.Santhiyakumari Professor Knowledge Institute Of Technology
Rvs College Of Engineering And
75 Professor
Dr V Gunaraj Technology
76 Professor
Dr S S Sivaraju Technology
77 Professor
Dr.M.Sundar Prakash Balaji Technology
Srm Institute Of Science And
78 Associate Professor
Dr C Subramani Technology
79 Dr Kavitha Murugesan Professor Vedavyasa Institute Of Technology
Srm Institute Of Science And
80 Professor
Boopathi Sengodan Technology
Kongu Engineering College
Nithyavathy N Perundurai Tamilnadu India
Dr.D.Vijendra Babu Technology
Dr.S.Sasikala Technology
Dr. L.K. Hema Technology
85 Dr.R.Josphineleela Associate Professor Panimalar Engineering College
86 Manju.M Associate Professor Panimalar Institute Of Technology
Alagumeenaakshi Muthiah Technology
Arun Kumar S Technology
Dr Maheswari K Technology
New Horizon College Of
Dr Vinoth Kumar K Engineering
Gunapriya K Engineering
K C R Nisha Engineering
xvii
ISBN: 979-8-3503-4821-7
NATIONAL REVIEWERS

1 Dr.N.Sureshkumar Professor VIT University
2 Dr.T.Ramkumar Professor VIT University
3 Dr. L.S. Jayashree Professor PSG College of Technology
4 Dr.M.Balamurugan Professor Arulmurugan College of engineering
5 Dr.S.P. Shantharajah Professor VIT University
6 Dr.S.Nandagopal Professor Nandha College of Technology
7 Dr.M.Vijayakumar Professor Nandha College of Technology
8 Dr.T.Padma Professor Sona College of Technology
9 Dr. K.Sankar Professor Sri Venkateswara College of Engineering
Dr. Newlin
10 Assistant Professor Anna University Regional Center
Rajkumar
11 Dr.Geetha Associate Professor Jeppiaar Mamallan Engg College
Madanapalle Institute of Technology and
12 Dr.Thamarai Professor
Science
13 Dr.S.Geetha Associate Professor Coimbatore Institute of Technology
14 Dr. S. Lovelyn Rose Professor PSG College of Technology
15 Dr.T.Senthilkumar Professor A.C.S. College of Engineering
16 Dr.N. Rakesh Professor Amrita Vishwa Vidyapeetham- University
17 Dr.G.M.Tamilselvan Professor Bannariamman Institute of Technology
18 Dr.K. Sakthisudan Assistant Professor Madanapalle Institute of Tech. & Science
19 Dr.Balamurugan Professor K.S.R College of Technology
20 Dr.D.Nirmal Associate Professor Karunya University
21 Dr.Rizwana Associate Professor KPRIT
22 Dr.Sudha Associate Professor MVJ College of Engineering
23 Dr. P.S. Periasamy Professor KSR College of Engineering
24 Dr. S. Smys Professor Karpagam College of Engineering
25 Dr.B.Arunadevi Associate Professor Dr. N.G.P. Institute of Technology
Dr.K.N.
26 Professor Dr. Mahalingam college of Engineering
Vijayakumar
Dr. Bhavani
27 Professor Karpagam University
Sridharan
28 Dr. V.V.Karthikeyan Professor Karpagam University
Dr. Michel Kingston
29 Associate Professor Dayanandha sager college of engineering
Raj
Dr. Raja
30 Assistant Professor SNS College of Technology
Ranganathan
xviii
ISBN: 979-8-3503-4821-7
31 Dr.K. Santhi Professor K S Rangasamy College of Technology

Dr. R. Jagadeesh
32 Professor VIT University Chennai Campus
Kannan
33 Dr. S.Karthik Professor SNS College of Technology
Hindusthan College of Engineering &
34 Dr. N. Sengottaiyan Professor
Technology
35 Dr.T.Senthilkumar Professor Amrita School of Engineering
36 Dr. T.Kalaikumaran Professor SNS College of Technology
37 Dr.A.R.Revathi Associate Professor SRM-Valliammai Engineering College
38 Dr.G.Rajesh Assistant Professor Anna university- MIT Campus
39 Dr. M. Mathivanan Professor ACS College of Engineering Bangalore
40 Dr.T.Rajendran Assistant Professor Arignar Anna Government College
41 Dr.K.Saravanan Associate Professor Kalasalingam University
42 Dr.M.Marikkannan Assistant Professor I.R.T.T
43 Dr.K.Sankar Professor MGM's College of Engineering
44 Dr.S. Rathi Assistant Professor Government College of Technology
45 Dr.G.R.Sakthidharan Assistant Professor Gokaraju Rangaraju Institute
46 Dr. K.Senthil Associate Professor Mohamed Sathak Engineering College
47 M.Parthiban Assistant Professor VSB Engineering College
48 Mr. T.Senthil Kumar Assistant Professor SRM University
49 Ms. G N BEENA Associate Professor Gokaraju Rangaraju Institute
50 Prof.R.Karthikeyan Assistant Professor Kilakarai
51 Dr. B. Muthu Kumar Professor Syed Ammal Engineering College
52 Balika J Chellaiah Assistant Professor SRM University
53 Mr. Dhanapal.R Project Consultant Vento Infotech India Private Limited
Gokaraju Rangaraju Institute of Engineering
54 Dr. B Sankara Babu Professor
and Technology
55 Dr.N.Karthikeyan Professor Syed Ammal Engineering College
Mr. Krishna
56 Professor Gokaraju Rangaraju Institute of Eng.
Chythanya Nagaraju
Malla Reddy Engineering College for
57 Dr.P.Manikandan Professor
Women
58 Mr. S.Karthick Assistant Professor Hindustan Institute of Technology
Jayamukki Institute of Technological
59 Dr.S.Kirubakaran Associate Professor
Services
GSSS Institute of Engineering &
60 Dr.P.Raviraj Professor
Technology for Women
Dr.V.Prasanna
61 Associate Professor Government College of Technology
Moorthy
xix
ISBN: 979-8-3503-4821-7
62 Dr. F. T. Josh Assistant Professor Karunya University

Dr.Mahalingam College of Engieering and
63 Dr. B.Vinoth Kumar Assistant Professor
Technology
Dr.NGP Institute of Engineering and
64 Dr.S.Jaganathan Professor
Technology
Arunachala college of engineering for
65 Dr.C.Agees Kumar Professor
women
66 Dr.C.Bharathiraja Associate Professor SRM University
67 Dr.Rajan Singaravel Assistant Professor PSG College of Technology
68 Dr.Sivarasu Professor Sri Eshwar College of Engineering
69 Dr.K.Prabakaran Professor KSR College of Engineering
70 Dr. N. Kumaresan Assistant Professor Anna University
71 Mr.M Sakthivel Team Lead Robert bosch
72 Dr.N R T Srinivasan Professor Amritha University
73 V A Saravanan Team Lead Wipro
74 Dr.Arun Vignesh Associate Professor GRIET
75 Mrs.S Aunchana Team Head Robert bosch
Mar Baselios Christian College of
76 Dr.P.Joby Professor
Engineering and Technology
Dr K
77 Associate Professor Adhi College of Engg & Technology
SAKTHISUDHAN
78 Dr R Anitha Associate Professor Adhi College of Engg & Technology
79 Dr.Saravanakumar N Professor Bannai amman Inatitute of Tech
80 Mr.I Harish R&D Engineer Vishli Industries private ltd
Enterprise Tech
81 Satheesh Kumar M Sonicwall Info Security Pvt Ltd
Supp Sr. Analyst
Dr.T K
82 Vice-Chairman Amritha University
Rameshkumar
83 Mr.A.Natarajan Assistant Professor PSGCT
Anna University Regional Centre
84 Dr.M.Yuvaraju Assistant Professor
Coimbatore
85 Dr.P.Rajendran Professor Knowledge Institute of Technology
86 Dr.V.Kumar Professor Knowledge Institute of Technology
JKKN College of Engineering &
87 Dr.C.Sureshkumar Professor
Technology
88 Dr.E.Govindaraj Professor MES College of Engineering
89 Dr.K.Sakthivel Professor KSR College of Technology
90 Dr. P.Karunakaran Professor Erode Sengunthar Engineering College
KPR Institute of Engineering and
91 Dr.M. Kathirvel Professor
Technology
92 Dr.S.Rajkumar Professor Info Institute of Technology
xx
ISBN: 979-8-3503-4821-7
Dr.A.
93 Professor KSR College of Technology
Gnanabaskaran
94 Dr.Sathish Kumar Professor Sri Ramakrishna Engineering College
Mutharasu
95 Technical Lead Wafer Space Semiconductor
Krishnamoorthi
96 Mr. Kirankumar Senior Developer Payoda technologies
97 Dr.R Anandhakumar Associate Professor Excel College of Technology
98 Dr.I.S.Akila Professor Coimbatore Institute of Technology
99 Dr.S Harikrishnan Professor Adhi College of Engg & Technology
Dr. Kotha Srinivasa
100 Assistant Professor Amrita University
Reddy
Kalasilingam Academy of Research and
101 Dr. P.Deepalakshmi Associate Professor
eduction
Dr.A.Roopa Malls Reddy College of Engineering and
102 Professor
Chandrika Technology
103 Dr.A.Chandrasekar Professor Malla reddy engineering college
Dr.B.Santhosh
104 Associate Professor GMRIT
Kumar
105 Dr.Kalidas Yeturu Assistant Professor IIT-Tiruati
106 Hirendra Hajare Assistant Professor Ballarur Institute of Technology
107 Lalita C. Nene Assistant Professor Veermata Jijabai Technological Institute
108 Dr. Anala .A.Pandit Associate Professor Veermata Jijabai Technological Institute
109 Dean Sasthra Deemed University
110 Dr.Manivannan.D Associate Professor Sasthra Deemed University
Senior Associate
111 Dr.Manikandan.G Sasthra Deemed University
Professor
Dr. J. AROKIA
112
RENJIT Professor & Head Jeppiaar Engineering College
113 Dr.P.JESU JAYARIN Associate Professor Jeppiaar Engineering College
114 Dr. Anita Kanavalli Professor Ramaiah Institute of Technology
115 Dr. Annapurna P. Patil Professor Ramaiah Institute of Technology
116 Dr.Subba Reddy N V Professor Manipal Institute of Technology
117 Dr. Preetham Kumar Professor Manipal Institute of Technology
118 Dr. G. P. Bhole Assistant Professor Veermata Jijabai Technological Institute
119 B. B. Meshram Professor Veermata Jijabai Technological Institute
120 Dr. R. D. Daruwala Professor & Head Veermata Jijabai Technological Institute
Dr. N
121
Ramasubramanian Professor National Institute of Technology
122 Dr. M.Bhaskar Professor National Institute of Technology
123 Dr.P.Anandhakumar Professor Anna university
xxi
ISBN: 979-8-3503-4821-7
124 Dr.M.R.Sumalatha Professor Anna university

Debotosh
125
Bhattacharjee Professor Jadavpur University
126 Sarbani Roy Professor Jadavpur University
127 Dr. Amirtharajan .R Professor Sastra Deemed University
128 Dr. Muthubalan .V Professor Sastra Deemed University
129 Dr. K. R. Baskaran Professor Kumaraguru College of Technology
130 Dr. Suganthi N Professor Kumaraguru College of Technology
Hindustan Institute of Technology and
131
Dr.C.V. Suresh Babu Professor Science
Dr. P.Selvi Hindustan Institute of Technology and
132
Rajendran Professor Science
133 Dr Palanisamy C Professor Bannari Amman Institute of Technology
134 Dr Nataraj R V Professor Bannari Amman Institute of Technology
135 Parvathi R M S Professor Sri Ramakrishna Institute of Technology
136 Suresh Kumar M Professor Sri Ramakrishna Institute of Technology
137 Dr.P.Anandhakumar Professor Madras Institute of Technology
138 Dr.M.R.Sumalatha Professor Madras Institute of Technology
139 Mr. K. Karthikeyan Assistant Professor SNS College of Engineering
140 Mr. E.P. Prakash Assistant Professor SNS College of Engineering
141 Dr B Niveditha Assistant Professor PSG College of Technology
142 Dr C Maheswari Assistant Professor PSG College of Technology
143 Dr.R.Divya Assistant Professor PSG College of Technology
144 Dr V Ranganayaki Assistant Professor Dr NGP Institute of Technology
145 Dr L.Ashok Kumar Professor PSG College of Technology
Dr D Karthika
146 Associate Professor PSG College of Technology
Renuga
147 Dr Pandian Gurusamy Professor Vishnu Institute of Technology
148 Dr .Supraja Assistant Professor SRM Institute of Science and Technology
149 Dr M Rajalakshmi Associate Professor Coimbatore Institute of Technology
150 Dr.R.Anuradha Associate Professor Sri Ramakrishna Engineerung College
Dr E Grace Mary
151 Associate Professor Karunya University
Kanaga
Dr.P Santhi
152 Associate Professor National Institute of Technology
Thilagam
153 Dr B Surendiran Associate Professor National Institute of Technology
154 Dr R Dhanalakshmi Associate Professor Indian Institute of Information Technology
155 Dr Jafar Ali Ibrahim Associate Professor QIS College of Engineering and Technology
xxii
ISBN: 979-8-3503-4821-7
156 J.Sheik Dawood Senior Manager Kryptos technologies

PSNA College of Engineering and
157 Dr T Hemalatha Professor
Technology
158 B M Prabhu Assistant Professor Park College of Engineering and Technology
159 V.Mani Assistant Professor Park College of Engineering and Technology
160 R Sundar Assistant Professor Dr NGP Institute of Technology
161 Jayasree.S Assistant Professor SNS College of Technology
KPR Institute of Engineering and
162 G. Saravanan
Assistant Professor Technology
163 Dr.N.Vinothkumar Assistant Professor Kumaraguru College of Technology
164 V.Ranjith Assistant Professor Park College of Engineering and Technology
165 R.Sureshkumar Assistant Professor Kumaraguru College of Technology
166 Dr.S.Sathishkumar Assistant Professor Jansons Institute of Technology
167 B.Gokulavasan Assistant Professo Sri Eshwar College of Engineering
168 Dr.Karunamoorthy Assistant Professor Kumaraguru College of Technology
169 Suresh Babu.V Professor College of Engineering Trivandrum
170 Biji Jacob Professor College of Engineering Trivandrum
171 Ajayan K.R Professor College of Engineering Trivandrum
172 Sindhu.N Associate Professor College of Engineering Trivandrum
173 Binu.L.S Professor College of Engineering Trivandrum
174 Sreelatha.G Professor College of Engineering Trivandrum
175 Suresh Kumar.E Professor College of Engineering Trivandrum
xxiii
INDEX
PAPER
S.NO ID TITLE PG.NO
A BLOCKCHAIN BASED CERTIFICATE
1 CS100 AUTHENTICATION SYSTEM 01
HAND GESTURES FOR MOUSE MOVEMENTS WITH
2 CS115 HAND AND COMPUTER INTERACTION MODEL 01
ARTIFICIAL INTELLIGENCE DEPLOYMENT AND THE
3 CS117 POTENCY OF HRM FUNCTIONS 01
THE APPROACH OF DIGITAL MARKETING IS BEING
REVOLUTIONIZED BY MACHINE LEARNING
4 CS120 POSSIBILITIES 02
NEW REAL TIME CLINICAL DECISION SYSTEM USING
5 CS126 MACHINE LEARNING – DISEASE PREDICTION 02
DIAGNOSIS OF LUNG AND PLEURAL DISEASES BY
6 CS128 MACHINE LEARNING ALGORITHMS 03
RECOGNIZING HANDWRITTEN OFFLINE TAMIL
7 CS133 CHARACTER USING VAE-GAN & CNN 03
BLOCKCHAIN APPLICATIONS IN HEALTHCARE-
8 CS134 RECENT ADVANCEMENTS 03
BLOCKCHAIN BASED HEALTHCARE SYSTEM TO
9 CS135 SECURE HEALTH RECORDS 04
A SYSTEMATIC INVESTIGATION OF CYBER
HARASSMENT INTENTION BEHAVIORS AND ITS
10 CS150 IMPACTS ON SOCIAL MEDIA PLATFORMS 04
STOCK PRICE PREDICTION USING RANDOM FOREST
11 CS152 CLASSIFIER AND BACKTESTINGCH 05
COMPARISON OF CRYPTOGRAPHY ALGORITHMS IN
12 CS155 BLOCKCHAIN 05
BITCOIN PRICE PREDICTION USING MACHINE
13 CS162 LEARNING 06
HUMAN ACTIVITY RECOGNITION FOR ANALYSING
14 CS163 FITNESS DATASET USING A FITNESS TRACKER. 06
ANALYSIS, SMARTPHONE, RANDOM FOREST,
15 CS164 MACHINE LEARNING, CLASSIFICATION, SENSORS 06
AN EFFECTIVE SENTIMENT ANALYSIS
CLASSIFICATION FOR OPINION PREDICTION USING
16 CS167 MACHINE LEARNING SCHEMES 07
SENTIMENT ANALYSIS, KNN, EUCLIDEAN
17 CS175 DISTANCE, MACHINE LEARNING AND OPINION. 07
COMPARATIVE STUDY OF DETECTION AND
ANALYSIS OF DIFFERENT MALWARE WITH THE
18 CS189 HELP OF DIFFERENT ALGORITHM 07
CHATBOT BASED VIRTUAL DIAGNOSIS ASSISTANT
FOR SKIN DISEASES AND BONE FRACTURE
19 CS197 DETECTION USING NEURAL NETWORK 08
20 CS199 SCALING BLOCKCHAIN SYSTEMS USING DAG 08
LANE COLORATION IN REAL TIME VEHICULAR AD-
21 CS202 HOC NETWORKS (VANETS) 08
CANNY EDGE DETECTION, KALMAN FILTER, LANE
COLORATION, DEEP LEARNING, INDIAN ROADS,
22 CS221 DBSCAN 09
23 CS226 IMPLEMENTATION OF YARA RULES IN ANDROID 09
AN ENSEMBLE OF MACHINE AND DEEP LEARNING
MODELS FOR REAL TIME CREDIT CARD SCAM
24 CS227 RECOGNITION 10
GRAPH THEORETIC APPROACH TO IOT
25 CS229 APPLICATIONS 10
APPLICATIONS OF BLOCKCHAIN TECHNOLOGY IN
MODERN EDUCATION SYSTEM – SYSTEMATIC
26 CS241 REVIEW 11
BOOKCHAIN: A SECURE LIBRARY BOOK STORING
AND SHARING IN ACADEMIC INSTITUTIONS USING
27 CS242 BLOCKCHAIN TECHNOLOGY 11
E- GOVERNENCE SERVICES USING ARTIFICIAL
28 CS243 INTELLIGENCE TECHNIQUES 11
BDDOS: BLOCKING DISTRIBUTED DENIAL OF
SERVICE FLOODING ATTACKS WITH DYNAMIC PATH
29 CS244 DETECTORS 12
30 CS247 ADAPTIVE TRAFFIC CONTROL SYSTEM USING YOLO 12
CLOUD COMPUTING BASED (SERVERLESS
COMPUTING) USING SERVERLESS ARCHITECTURE
FOR DYNAMIC WEB HOSTING AND COST
31 CS259 OPTIMIZATION 13
OBSTACLE DETECTION AND NAVIGATION SUPPORT
USING SMART SHOE FOR VISUALLY IMPAIRED
32 CS260 PERSON 13
33 CS261 ONLINE MEDICAL SYSTEM USING ELGAMAL 13
PERSONALIZED SECURE E-IDENTITY CARD USING
34 CS265 BLOCK CHAIN 14
AUTOMATED YOUTUBE VIDEO TRANSCRIPTION TO
SUMMARIZED TEXT USING NATURAL LANGUAGE
35 CS266 PROCESSING 14
PHISHING WEBSITES DETECTION USING MACHINE
CRYPTANALYSIS OF AN ANONYMOUS AND
TRACEABLE GROUP DATA SHARING IN CLOUD
37 CS268 COMPUTING 15
VISUAL CRYPTOGRAPHY FOR SECURE BANKING
38 CS269 APPLICATION 15
BANKING DATA MIGRATION FROM ON-PREMISE TO
39 CS271 CLOUD 16
TEXT RECOGNITION BASED ON ENCODER AND
40 CS272 DECODER FRAMEWORK 16
AN EFFECTIVE APPROACH FOR PREPARATION OF
NEWS SUMMARY USING WEB SCRAPING
41 CS273 TECHNIQUES 17
DETECTION OF CYBER ATTACK USING ARTIFICIAL
42 CS280 INTELLIGENCE 17
USING HONEY POT TECHNIQUE RANSOMEWARE
43 CS281 GET DETECTED 17
COMPREHENSIVE SURVEY ON INTRUSION
DETECTION SYSTEM USING DEEP LEARNING
AUTHENTICATION SYSTEM FOR UNIVERSITY
45 CS284 ADMINISTRATIVE ACTIVITIES 18
ENHANCING THE SECURITY FEATURES IN REMOTE
KEYLESS ENTRY SYSTEM RAGAVENDRA PRASAD M,
46 CS285 SUJATHA G 19
47 CS286 CLOUD STORAGE SECURITY MODEL 19
48 CS290 ECC-BASED AUTHENTICATION FOR SECURE TDMA 19
SENTIMENT ANALYSIS OF TWITTER REVIEWS USING
49 CS293 SVM, ANN AND CNN MODELS 20
A PECULIAR IMAGE ENCRYPTION TECHNIQUE FOR
50 CS302 MOBILE 20
DEEP LEARNING-BASED SENTIMENT ANALYSIS ON
51 CS303 TAMIL TEXTS IN ROMAN SCRIPT 21
YOLO: HUMAN DETECTION-BASED INTELLIGENT
52 CS304 HOME AUTOMATION USING IOT 21
DETECTION OF CYBERBULLYING BASED ON ONLINE
53 CS306 SOCIAL NETWORKS 21
CLOSEST CELESTIAL BODY SEARCH USING KD
54 CS317 TREES 22
TRUSTWORTHY WEB SERVICE-BUSINESS
ACTIVITIES (WSBA) USING AN EFFICIENT
55 CS320 BYZANTINE FAULT TOLERANCE ALGORITHM 22
INDUSTRY 4.0 DIGITAL TECHNOLOGIES AND
INFORMATION SYSTEMS: IMPLICATIONS FOR
MANUFACTURING FIRMS INNOVATION
56 CS321 PERFORMANCE 23
CLASSIFICATION OF INFLUENZA FROM CHEST
RADIOGRAPHY USING DEEP NEURAL
57 CS322 ARCHITECTURE 23
MULTIMODAL DEEP NEURAL NETWORK FOR
58 CS327 DIABETIC RETINOPATHY DETECTION 23
DESIGN OF SECURE QR PAYMENT SYSTEM USING
59 CS329 VISUAL CRYPTOGRAPHY METHOD 24
CONVOLUTION NEURAL NETWORKS BASED LUNGS
DISEASE DETECTION AND SEVERITY
60 CS332 CLASSIFICATION 24
OPTIMIZING VIRTUAL MACHINE DOWNTIME IN LIVE
MIGRATION USING THE LEAST FREQUENTLY USED
61 CS336 SERVICE-BASED PRE-COPY ALGORITHM 25
AGE AND GENDER PREDICTION USING ADAPTIVE
GAMMA CORRECTION AND CONVOLUTIONAL
62 CS339 NEURAL NETWORK 25
GAIT SPEED BASED INDIVIDUAL RECOGNITION
MODEL USING DEEP 2-D CONVOLUTIONAL NEURAL
63 CS345 NETWORK 26
E-COATING ULTRAFILTRATION SYSTEM
MAINTENANCE USING MACHINE LEARNING
UNSUPERVISED ONLINE VIDEO OBJECT
65 CS349 SEGMENTATION 26
A SYSTEMATIC REVIEW ON CLASSIFICATION OF
CYBER ATTACKS AND ITS PREVENTION
66 CS351 TECHNIQUES TO IMPROVE CYBER SECURITY 27
ANTI-ACCIDENT MECHANISM TO DETECT DRIVER
DROWSINESS INTEGRATED WITH ALERTING
67 CS354 MECHANISM: AN IOT-BASED MODEL 27
CLASSIFICATION OF TWEETS USING NATURAL
68 CS360 LANGUAGE PROCESSING FROM TWITTER API DATA 27
AUTOMATION OF BIG MART DATA SALES USING
69 CS363 SUPERVISED MACHINE LEARNING TECHNIQUES 28
ASPECT BASED SENTIMENT AWARE WORD
EMBEDDING FOR CROSS DOMAIN SENTIMENT
70 CS364 ANALYSIS 28
THE ROLE OF ERNIE MODEL IN ANALYZING HOTEL
71 CS367 REVIEWS USING CHINESE SENTIMENT ANALYSIS 29
A REVIEW OF BLOCKCHAIN TECHNOLOGY IN
SMART APPLICATIONS: RESEARCH ISSUES AND
72 CS369 OPEN CHALLENGES 29
EMPIRICAL INVESTIGATION ON THE SECURITY AND
73 CS370 THREAT MITIGATION OF CAMPUS SWITCHES 29
CLASSICAL REVIEW ON STOCK PROGNOSIS USING
74 CS376 LONG SHORT TERM MEMORY 30
AUTOMATED E-COMMERCE PRICE COMPARISON
WEBSITE USING PHP, XAMPP, MONGODB, DJANGO,
75 CS382 AND WEB SCRAPPING 30
ENHANCED DATA SECURITY USING TRUST
ATTRIBUTE-BASED ENCRYPTION IN CLOUD
76 CS383 COMPUTING 31
EGMPIP: ENHANCED GEOGRAPHIC MULTIPATH
ROUTING USING GOSSIP-BASED DIRECT NEIGHBOR
DISCOVERY ALGORITHM IN WIRELESS SENSOR
77 CS384 NETWORKS 31
EMERGENCY TRACKING SYSTEM USING
78 CS390 INTELLIGENT AGENT 32
INTELLIGENT SYSTEM; ARTIFICIAL INTELLIGENCE;
SMART WATCH; GPS/GSM; SMART PHONE
79 CS391 APPLICATION; MACHINE LEARNING 32
WEATHER PREDICTION USING LINEAR REGRESSION
80 CS392 MODEL 32
81 CS393 SEWAGE MONITORING SYSTEM USING IOT 33
CLASSIFICATION OF STATUS OF FRUITS USING DEEP
82 CS395 LEARNING ALGORITHMS 33
IMBALANCED BREAST CANCER PREDICTION USING
83 CS398 CNN THROUGH MULTI-NET FRAMEWORK 34
CLASSIFICATION OF BREAST CANCER DIAGNOSIS
84 CS403 USING MACHINE LEARNING ALGORITHMS 34
LIFE-SAVING OPULENT AMBULANCE TRACKING
85 CS407 SYSTEM 34
AI BASED SPEECH RECOGNITION OF LITERACY TO
86 CS408 IMPROVE TRIBAL ENGLISH KNOWLEDGE 35
A CNN BASED MACHINE LEARNING SCHEME TO
DETECT LUNG CANCER DETECTION FROM CT SCAN
87 CS411 IMAGES 35
COMPARATIVE ANALYSIS OF CANNY EDGE
DETECTION-BASED SMART TRAFFIC CONTROL
88 CS412 SYSTEM USING MULTI-TYPE IMAGE PROCESSING 35
A COMPARATIVE ANALYSIS OF RETINAL DISEASE
IMAGE CLASSIFICATION FOR OCT USING DEEP
89 CS413 LEARNING TECHNIQUES 36
IMAGE CAPTIONING FOR INFORMATION
90 CS418 GENERATION 36
AN IMPLEMENTATION OF BLOCKCHAIN
TECHNOLOGY IN COMBINATION WITH IPFS FOR
91 CS425 CRIME EVIDENCE MANAGEMENT SYSTEM 36
METHODS FOR AUTOMATICALLY IDENTIFYING AND
92 CS426 DIAGNOSING MENTAL HEALTH ISSUES 37
MENTAL HEALTH DETECTION AND ANTICIPATION
93 CS427 ANALYSIS USING DEEP LEARNING 37
AUTOMATED MECHANISM FOR IMPROVING
94 CS428 PRECISION IRRIGATION USING MACHINE LEARNING 38
VOICE CONTROLLED HOME AUTOMATION SYSTEM
95 CS430 USING GOOGLE ASSISTANTS 38
REMOTE HEALTH PREDICTION SYSTEM USING
FACE MASK DETECTION USING SUPPORT VECTOR
97 CS432 MACHINE 39
DEFECT EXPOSURE IN VEGETABLES AND FRUITS
98 CS433 USING MACHINE LEARNING 39
AUTOMATED ANSWER SCRIPT EVALUATION USING
99 CS438 DEEP LEARNING 40
COMPARATIVE STUDY OF QUANTUM MACHINE
LEARNING ENHANCED SVM AND CLASSICAL SVM
100 CS439 FOR BRAIN STROKE PREDICTION 40
ENHANCING THE SECURITY FOR HEALTHCARE
101 CS440 DATA USING BLOCKCHAIN TECHNOLOGY 40
SENTIMENTAL ANALYSIS RECOGNITION IN
102 CS442 CUSTOMER REVIEW USING NOVEL-CNN 41
A NOVEL TWO-WAY CROSS-TAB MACHINE
LEARNING APPROACH FOR PREDICTING LIFE
INSURANCE USING BIVARIATE EXPLORATORY
SHOPIFY – SHOP AUGMENTATION AND
104 CS445 RECOMMENDATION SYSTEM 42
HEART DISEASE PREDICTION USING MACHINE
PITCH EXTRACTION AND NOTES GENERATION
106 CS449 IMPLEMENTATION USING TENSORFLOW 43
RFID-ENABLED AUTOMATED BILLING SYSTEM FOR
107 CS451 RETAIL ESTABLISHMENTS 43
PRIVATE CLOUD STORAGE USING RASPBERRY PI
VIA VIRTUAL NETWORK COMPUTING - AN
CHEST DISEASE CLASSIFICATION USING
CONVOLUTIONAL NEURAL NETWORKSCHEST
DISEASE CLASSIFICATION USING CONVOLUTIONAL
109 CS453 NEURAL NETWORKS 44
110 CS454 WOMEN’S SAFETY WRISTBAND USING IOT 44
DESIGN AND IMPLEMENTATION OF ALZHEIMER'S
111 CS455 DISEASE DETECTION USING CGAN AND CNN 45
A DEEP LEARNING APPROACH BASED ON CT
IMAGES FOR AN AUTOMATIC DETECTION OF
112 CS456 POLYCYSTIC KIDNEY DISEASE 45
RECOGNITION OF TAMIL HANDWRITTEN
113 CS457 CHARACTERS USING SCRABBLE GAN 46
114 CS461 SMART DUSTBIN USING NODEMCU 46
IMPROVED NETWORK SECURITY BY UTILIZING
CLOUDY AUTOMATED WATERMARKING
THE EARLY DETECTION OF DEMENTIA DISEASE
116 CS471 USING MACHINE LEARNING APPROACH 47
KNOWLEDGE BASED PERCEPTIONS OF CHATBOTS IN
117 CS473 HEALTHCARE 48
YOLO BASED LICENSE PLATE DETECTION OF TRIPLE
118 CS474 RIDERS AND VIOLATORS 48
FAKE NEWS DETECTION USING MACHINE LEARNING
119 CS475 APPROACH 48
FAAP -A FARMER’S AMIABLE APPLICATION TO
ENHANCE AGRICULTURE USING MOBILE
120 CS476 TECHNOLOGY 49
CERVICAL CANCER DETECTION USING THE
121 CS477 ENSEMBLE METHOD IN MACHINE LEARNING 49
SUICIDAL THOUGHTS OF RECENT GENERATIONS- A
122 CS478 REVIEW 50
EMOTION PREDICTION THROUGH FACIAL
RECOGNITION USING MACHINE LEARNING: A
123 CS479 SURVEY 50
SMART DIAGNOSTIC SYSTEM FOR EARLY
DETECTION AND PREDICTION OF POLYCYSTIC
124 CS481 OVARY SYNDROME 50
AN APPROACH TO DETECT MELANOMA SKIN
125 CS482 CANCER USING FAST AI CNN MODELS 51
PREDICTION OF TYPHOID USING MACHINE
126 CS484 LEARNING AND ANN PRIOR TO CLINICAL TEST 51
TEXT DETECTION AND LANGUAGE IDENTIFICATION
127 CS485 IN NATURAL SCENE IMAGES USING YOLOV5 52
THRESHOLD BASED FEATURE SELECTION FOR
128 CS488 ANEMIA PREDICTION 52
RECURRENT GASTRIC CANCER PREDICTION USING
129 CS489 RANDOMIZED SEARCH CV OPTIMIZER 53
DATA SECURITY AND AUTHENTICATION IN
130 CS490 MODERN APPLICATIONS AND CLOUD COMPUTING 53
A NOVEL METHOD FOR PNEUMOTHORAX
DIAGNOSIS AND SEGMENTATION USING DEEP
131 CS495 CONVOLUTION NEURAL NETWORK 54
TWITTER SENTIMENT ANALYSIS USING
COLLABORATIVE MULTI LAYER
132 CS502 PERCEPTRON(CMLP) CLASSIFIER 54
PREDICTION OF PARKINSON’SDISEASEUSING
MACHINE LEARNING BASED ON VOCAL
133 CS504 FREQUENCY 54
MACHINE LEARNING-BASED CROP
RECOMMENDATIONS FOR PRECISION FARMING TO
134 CS505 MAXIMIZE CROP YIELDS 55
SALES PREDICTION FOR FOOD PRODUCTS USING
135 CS506 MACHINE LEARNING 55
A PERLUSTRATION ON THE OPTIMISTIC PROGNOSIS
OF CHRONIC RENAL FAILURE USING
EVOLUTIONARY MACHINE LEARNING AND DEEP
WHALE SPECIES CLASSIFIER USING DORSAL FIN
137 CS510 AND LATERAL BODY CHARACTERISTIC 56
HEART DISEASE PREDICTION-BASED ON
CONVENTIONAL KNN AND TUNED-HYPER
138 CS511 PARAMETERS OF KNN: AN INSIGHT 57
ALIVE HUMAN DETECTION IN WAR FIELDS AND
EARTHQUAKES USING PIR SENSOR AND
139 CS512 MONITORING THE HEARTBEAT OF THE VICTIMS 57
PERFORMANCE ANALYSIS OF BLOCKCHAIN
TECHNOLOGY USING INTER PLANETARY FILE
SYSTEM (IPFS) DATABASE IN COMPARISON WITH
140 CS514 BLOCK CHAIN DATABASE 57
SYSTEMATIC HARMFUL SIGNS DETECTION FOR
141 CS515 WOMEN’S SAFETY USING NEURAL NETWORKS 58
CLASSIFICATION OF SATELLITE AGRICULTURE
IMAGERY TO IDENTIFY CROP TYPE USING DEEP
142 CS516 LEARNING TECHNIQUES - A REVIEW 58
SYSTEM FOR DETECTING INTRUSIONS USING
143 CS519 RASPBERRY PI 59
MULTI-CLASS CLASSIFICATION OF INSECTS USING
144 CS520 DEEP NEURAL NETWORKS 59
COPY-MOVE IMAGE FORGERY DETECTION USING
MULTI-OBJECT DETECTION AND TRACKING USING
AN REAL-TIME OBJECT DETECTION METHOD FOR
THE VISUALLY IMPAIRED USING MACHINE
DETECTION OF ANIMALS FOR ROAD SAFETY USING
148 CS526 DEEP LEARNING METHOD 61
AUTOMATIC IMAGE CAPTIONING USING CNN AND
149 CS529 LSTM 61
ANIMAL DETECTION, CLASSIFICATION, DEEP
LEARNING ALGORITHMS, TEMPLATE MATCHING
150 CS530 ALGORITHM. 62
EFFECTIVE MODEL FOR IMPROVING SYMPTOMS-
BASED DISEASE PREDICTION BY BIMM - BERT
151 CS532 ALGORITHM 62
COMPARATIVE ANALYSIS OF DIFFERENT
APPROACHES TO THE MUSIC RECOMMENDATION
152 CS536 SYSTEM 62
RECOMMENDATION SYSTEM, MUSIC
RECOMMENDATION SYSTEM, CONTENT-BASED
FILTERING, COLLABORATIVE FILTERING, HYBRID
153 CS538 FILTERING, SESSION-BASED 63
PREDICTION OF RENTAL DEMANDS USING
154 CS540 DIFFERENT MACHINE LEARNING ALGORITHMS 63
DETECTION OF PERSONAL ATTACKS IN MOBBING
ON SOCIAL MEDIA USING DEEP LEARNING
EXPLORING THE POTENTIAL OF BLOCKCHAIN
TECHNOLOGY FOR SECURING AND MANAGING
156 CS543 STUDENT 64
GLAUCOMA PREDICTION USING SUPERVISED
AUTOMATED STOCK PRICE PREDICTION USING
158 CS545 LSTM-ANN 65
COMPARING AND DESCRIBING THE
CHARACTERISTICS OF SWARM OPTIMIZATION IN
159 CS546 BIO-INSPIRED 65
IOTSDL: INTERNET OF THINGS SECURITY FOR DEEP
160 CS548 LEARNING TECHNIQUES-A 66
PREDICTION OF WATER QUALITY PARAMETERS BY
161 CS550 IOT AND MACHINE LEARNING 66
FUNCTIONAL OVERVIEW OF INTEGRATION OF AIML
162 CS551 WITH 5G AND BEYOND THE NETWORK 67
A MACHINE LEARNING APPROACH FOR PREDICTING
163 CS554 CROP SEASONAL YIELD AND COST FOR SMART 67
ONE-DIMENSIONAL CONVOLUTIONAL NEURAL
NETWORK AND LONG SHORT-TERM MEMORY
164 CS557 CONVOLUTIONAL 68
E-COMMERCE CUSTOMER CHURN PREDICTION
SCHEME BASED ON CUSTOMER BEHAVIOUR USING
165 CS558 MACHINE 68
SMART HOME AUTOMATION SECURITY USING
166 CS560 TYPE-2 FUZZY-BASED APPROACH THROUGH IOT 69
CLASSIFYING THE STATUS OF CKD USING
RANDOMIZED WEIGHTED OPTIMIZATIONMODEL
167 CS561 USING 69
FUNDUSDBNET: METHODOLOGY TO DETECT
168 CS562 RETINAL FUNDUS DISEASES 70
CERVICAL CELL CLASSIFICATION USING DEEP
APPLICATION OF LATEST AUTHENTICATION
170 CS567 TECHNOLOGIES IN VEHICULAR CLOUD 70
ADAPTIVE ENERGY ROUTING PROTOCOL USING
SPIDER OPTIMIZATION IN WIRELESS SENSOR
171 CS568 NETWORKS 71
A SURVEY ON EXISTING SYSTEM DESIGN USED FOR
MANAGING AMBULANCE BOOKING THROUGH
172 CS569 MOBILE APP 71
A REVIEW ON TECHNIQUES USED FOR FACE
AUTHENTICATION BASED SMARTDOOR BELL
173 CS570 SYSTEM USING IOT 72
EFFECT OF SUPERVISED MACHINE LEARNING
174 CS571 ALGORITHMS ON RICE PLANT DISEASES 72
FARMER’S FRIENDLY ROBOT USING INTERNET OF
175 CS575 THINGS 72
AN IMPROVED SYSTEM FOR STUDENTS FEEDBACK
ANALYSIS USING SUPERVISED PROBABILITY
HAND SIGN CLASSIFICATION TECHNIQUES USING
SUPERVISED AND UNSUPERVISED LEARNING
177 CS579 ALGORITHMS 73
RELATING THE DT FRAMEWORK IN SECURING THE
178 CS583 HEALTHCARE DATA USING BLOCKCHAIN 74
V-VERIFY WITH BIOMETRIC SYSTEM USING DESIGN
179 CS585 THINKING FRAMEWORK 74
AI - UNDERWATER DRONE IN PROTECTION OF
WATERWAYS BY RELATING DESIGN THINKING
180 CS588 FRAMEWORK 75
NEURAL NETWORK-BASED SOFTWARE
VULNERABILITY CLASSIFICATION TO ENHANCE
181 CS589 ACCURACY 75
PREDICTION OF MANGO LEAF DISEASES USING
182 CS591 CONVOLUTIONAL NEURAL NETWORK 75
RECOGNITION AND PROCESSING OF PHISHING
183 CS593 EMAILS USING NLP: A SURVEY 76
ADVANCMENT IN DRIVER ASSISTANCE SYSTEM
184 CS599 USING IOT TO PREVENT ACCIDENTS 76
COMPARATIVE PERFORMANCE ASSESSMENT OF
MACHINE LEARNING ALGORITHMS TO PREDICT
185 CS601 CARDIOVASCULAR DISEASE 77
A VISIONARY APPROACH TO DETECT SPOOFING
186 CS603 WEBSITE USING MACHINE LEARNING ALGORITHMS 77
ARTIFICIAL INTELLIGENCE-BASED HEALTH CARE

MONITORING AND RECORD STORAGE SYSTEM
187 CS605 77
PREDICTION OF BRAIN TUMOR CLASSIFICATION BY
188 CS606 USING CNN 78
AUTOMATIC VIDEO SUMMARIZATION AND
189 CS608 CLASSIFICATION BY CNN MODEL: DEEP LEARNING 78
BRAINY BIN FOR METROPOLITAN CITIES USING
190 CS619 DESIGN THINKING METHODOLOGY 79
IMPROVING EFFICIENCY OF GARBAGE COLLECTION
191 CS621 VEHICLE FOR SMART CITIES 79
SOCIALIZING PLATFORM FOR EDUCATIONAL
192 CS623 INSTITUTIONS 80
SOFT ATTENTION BASED DENSENET FOR SEVERITY
193 CS624 GRADING OF DIABETIC MACULAR EDEMA 80
ANALYZE THE POLITICAL PREFERENCE OF A
COMMON MAN BY USING DATA MINING AND
EFFICIENT GESTURE-BASED PRESENTATION
CONTROLLER USING TRANSFER LEARNING
195 CS628 ALGORITHM 81
SAFE TRADE – A STOCK RECOMMENDER USING
AN EXPLORATION OF EVALUATING THE
PERFORMANCE OF MALWARE DETECTION IN THE
197 CS634 CLOUD ENVIRONMENT 82
AN IMAGE BASED CLASSIFICATION AND
PREDICTION OF DISEASES ON COTTON LEAVES
198 CS635 USING DEEP 82
199 CS638 HYPERAUTOMATION-BEYOND RPA: 83
MUSIC RECOMMENDATION BASED ON FACIAL
200 CS639 EXPRESSIONS AND MOOD DETECTION USING CNN 83
ANALYSES OF MACHINE LEARNING TECHNIQUES
FOR SIGN LANGUAGE TO TEXT CONVERSION FOR
201 CS640 SPEECH IMPAIRED 84
SECURING CLOUD IDENTITY INFORMATION UNDER
202 CS641 KEY INTRODUCTION 84
CLASSIFICATION OF FOODS BASED ON
203 CS644 INGREDIENTS 85
204 CS645 PITCH AND TIMBRE ANALYSIS IN ACOUSTICS 85
ANALYSIS OF DYNAMIC SCHEDULING FOR EDGE
205 CS648 CLOUD COMPUTING 86
AN APPROACH FOR CERVICAL AND BREAST
CANCER CLASSIFICATION USING DEEP LEARNING: A
206 CS650 COMPREHENSIVE SURVEY 86
A SURVEY ON EXPLORING THE EFFECTIVENESS OF
207 CS651 IOT ASED HOME SECURITY SYSTEMS 86
SOCIAL MEDIA NEWS CLASSIFICATION USING
SOCIAL MEDIA NEWS CLASSIFICATION USING
EXTENSIVE REVIEW ON PREDICTING HEART
DISEASE USING MACHINE LEARNING AND DEEP
AN EFFECTIVE ARCHITECTURE FOR IMAGE
211 CS658 COLORIZATION USING ADAPTIVE GAN 88
RECOGNITION AND CLASSIFICATION OF IMAGE
BASED COAL AND COAL GANGUE USING DEEP
CS659 LEARNING TECHNIQUES 88
CONVERSION OF ECG SIGNALS AND HEART DISEASE
CS660 DETECTION USING MACHINE LEARNING
212 ALGORITHM 89
FORECASTING THE SOLAR POWER WITH
CS661 DIFFERENTIATION OF DATA FIGURES USING
213 NEURAL NETWORKS 89
MALWARE PREDICTION AND CLASSIFICATION FOR
CS662 ANDROID APPLICATIONS USING MACHINE
214 LEARNING TECHNIQUES 90
MULTICLASS SKIN CANCER IMAGE CLASSIFICATION
CS664
215 USING CONVOLUNTIONAL NEURAL NETWORK 90
ANALYSIS OF IRREGULAR MENSTRUAL CYCLE AND
CS665
216 ITS CONSEQUENCES 90
ACMAPPING: AGRICULTURAL CROP
CS666 MAPPING/RETRIEVAL USING SYNTHETIC APERTURE
217 RADAR (SAR) DATA 91
A SURVEY ON TEXT SUMMARIZATION IN URDU
CS668 LANGUAGE USING MACHINE LEARNING
218 TECHNIQUES 91
EFFICIENT ROAD STRUCTURAL DESIGN AND
CS675 TRAFFIC ACCIDENT ANALYSIS USING SUPERVISED
219 LEARNING ALGORITHMS 92
220 CS678 AGROCULTURE WEB APPLICATION 92
A COMPARATIVE ANALYSIS BETWEEN SINGLE &
CS680 DUAL-HANDED BANGLADESHI SIGN LANGUAGE
221 DETECTION USING CNN BASED APPROACH 93
BIG DATA ANALYTICS AND MACHINE LEARNING
CS684
222 TECHNIQUES TO MANAGE THE SMART GRID 93
DRUG CONSUMPTION DETECTION BY EYE USING
CS688
223 IMAGE PROCESSING TECHIQUES 94
SMART MANUFACTURING IN INDUSTRIES USING
CS689
224 INTERNET OF THINGS 94
ASTHMA DISEASE RISK PREDICTION USING
CS690
225 MACHINE LEARNING TECHNIQUES 95
A STRUCTURED ANALYSIS ON IPL 2022 MATCHES BY
CS691 APPROACHING VARIOUS DATA VISUALIZATION
226 AND ANALYTICS 95
DEVELOPMENT OF TEXT CHARACTERIZATION
CS692
227 USING NATURAL LANGUAGE PROCESSING 95
SECURE EXAMINATION SYSTEM WITH YOLOV3 &
CS697
228 DLIB 96
HATE SPEECH ANALYSIS USING SUPERVISED
CS703
DATA ANALYTICS FOR PHISHING ATTACK
CS704
230 DETECTION USING DEEP LEARNING TECHNIQUES 97
INFORMATION EXTRACTION FROM TEXT MESSAGES
CS705
231 USING NATURAL LANGUAGE PROCESSING 97
A SURVEY PAPER ON WEED IDENTIFICATION USING
CS707
232 DEEP LEARNING TECHNIQUES 97
A NOVEL APPROACH FOR MONITORING
AGRICULTURAL PRODUCTION PROCESS USING
CS711
WIRELESS SENSOR NETWORKS AND MACHINE
233 LEARNING 98
PERFORMANCE EVALUATION OF TRANSFER
CS715 LEARNING BASED MODELS ON SKIN DISEASE
234 CLASSIFICATION 98
TEXT ANALYSIS ON TWITTER DATA USING LSA AND
CS716
235 LDA 99
EFFICIENT LOAD BALANCING FOR MICRO SERVICES
CS717
236 BASED APPLICATIONS 99
CROSS SITE SCRIPTING (XSS) VULNERABILITY
CS726 DETECTION USING MACHINE LEARNING AND
237 STATISTICAL ANALYSIS 100
A SURVEY ON ANALYZING SOCIAL MEDIA DATA
CS731 AND THE DATA FORM PAST FOR EMERGENCY
238 MANAGEMENT 100
PRODUCT RECOMMENDATION SYSTEM USING DEEP
CS743
239 LEARNING BASED RECURRENT NEURAL NETWORK 101
PREDICTION OF CROPS BASED ON A MACHINE
CS747
240 LEARNING ALGORITHM 101
241 CS750 MOVIE RECOMMENDATION SYSTEM 102
ANALYSIS OF PERFORMANCE SCALABILITY ,
CS751 AVAILABILITY AND SECURITY IN DIFFERENT
242 CLOUD ENVIRONMENTS FOR CLOUD COMPUTING 102
SENTIMENT ANALYSIS OF BOOKS REVIEW USING
CS759
243 MACHINE LEARNING 103
ANAEMIA ESTIMATION FOR PATIENTS USING A
CS760
244 MACHINE LEARNING MODEL 103
UNIFROL - VERSATILE ONLINE CAMPUS
CS761 NETWORKING PLATFORM FOR STUDENTS USING
245 SOCIAL MEDIA ARCHITECTURE 104
CHATBOT FOR HEALTHCARE USING MACHINE
CS762
246 LEARNING 104
MACHINE LEARNING BASED INTELLIGENT CAREER
CS765
247 COUNSELLING CHATBOT (ICCC) 105
A REVIEW ON LONG SHORT TERM MEMORY BASED
CS766
248 STOCK PRICE PREDICTION TECHNIQUES 105
BIG DATA – ENHANCEMENT IN SPORTS : EXPLORING
CS769
249 THE IMPACT OF TECHNOLOGY INTEGRATION 105
IMAGE CAPTIONING USING DEEP LEARNING
CS774
250 TECHNIQUES FOR PARTIALLY IMPAIRED PEOPLE 106
ARTIFICIAL INTELLIGENCE APPLICATION FOR
CS775 EFFECTIVE CUSTOMER RELATIONSHIP
251 MANAGEMENT 106
AUDIO SUMMARIZATION OF PODCASTS IN REAL-
CS786
252 TIME 107
SMART AUTONOMOUS VOICE CONTROL OBSTACLE
CS789
253 AVOIDING ROBOT 107
AN ENHANCED YOLO BASED TRAFFIC ANALYSIS
CS799
254 SYSTEM 108
DISTRIBUTED INFERENCE APPROACH ON MASSIVE
CS800
255 DATASETS USING MAPREDUCE 108
CONTINUOUS DIGITAL SYSTEM ANALYSIS IN
DIFFERENT SYSTEM SOFTWARES USING
CS803
KEYLOGGERS TO VALIDATE THE NEED OF
256 SECURITY 109
FEVER DETECTION SYSTEM USING MACHINE
CS813
257 LEARNING ALGORITHM 109
MAXIMISING EFFICIENCY OF NETWORK RESOURCES
CS815 IN SDN THROUGH OPTIMAL PACKET ROUTING
258 USING REINFORCEMENT LEARNING 110
259 CS817 SECURE THE SMART GRID BY MACHINE LEARNING 110
ASSESSMENT BASED ON IOT FOR EFFICIENT
CS818 INFORMATION SURVEILLANCE REGARDING
260 HARMFUL STRIKES UPON FINANCIAL COLLECTION 111
AN IN-DEPTH ANALYSIS OF THE ROLE THAT ML AND
CS819 BIG DATA PLAY IN DRIVING DIGITAL MARKETING'S
261 PARADIGM SHIFT 111
AN ELABORATE REVIEW OF DIAGNOSING
CS821 GLAUCOMA DISEASE USING MACHINE LEARNING
262 AND DEEP LEARNING TECHNIQUES 112
ANALYSES OF FACE MASK DETECTION USING DEEP
CS822
263 NEURAL 112
MOVIE RECOMMENDATION SYSTEM USING
CS825 CONTENT BASED NOVEL FEATURE EXTRACTION
264 INFORMATION 112
265 CS827 YOUTUBE COMMENT SPAM DETECTION 113
PREDICTING USER CLICK BEHAVIOR ON SOCIAL
CS830
266 MEDIA ADS USING MACHINE LEARNING 113
A TECHNIQUE FOR DIGITAL IMAGE
CS834 WATERMARKING THAT DETECTS AND RESTORES
267 TAMPERING 116
AUTOMATIC EXTRACTIVE TEXT SUMMARIZATION
CS836
268 OF HINDI TEXT USING DEEP LEARNING APPROACH 116
AUTOMATIC CORRECTION OF TEXT USING
CS838
PROBABILISTIC ERROR APPROACH 115
COVID-19 POST-VACCINATION ADVERSE EFFECTS
CS839 PREDICTION WITH SUPERVISED MACHINE
269 LEARNING MODELS 115
270 CS840 AUTOMATIC EGG INCUBATOR USING IOT 115
COVID-19 DETECTION USING CNN MODEL WITH CT
CS841
271 SCAN IMAGES 116
ARTIFICIAL SPECIAL VISUAL GEOMETRY GROUP-
16(VGG) LEARNING MODEL FOR ANALYSING
CS842
ACCURACY AND PRECISION OF SARS-COV-2
272 FORECASTING 116
ANALYSIS OF CLOUD COMPUTING SECURITY
CS845 CHALLENGES AND THREATS FOR RESOLVING DATA
273 BREACH ISSUES 117
WEB INFORMATION RETRIEVAL USING VECTOR
CS846
274 SPACE MODEL AND DOCU-TALLY METRIC 117
BLOCKCHAIN ASSISTED CLOUD STORAGE FOR
CS849
275 ELECTRONIC HEALTH RECORDS 117
EXPLORING VARIOUS ASPECTS IN DIAGNOSING
CS850
276 AUTISM SPECTRUM DISORDER (ASD) 118
IDENTIFICATION OF LEAF DISORDERS USING
CS852
COMPARATIVE ANALYSIS FOR PREDICTION AND
CLASSIFICATION OF MENTAL HEALTH ISSUES AND
CS853
CHALLENGES USING HYBRID LEARNING
278 TECHNIQUES 119
279 CS857 SQUAD HELP – AN EMOTION BASED MUSIC PLAYER 120
SOLDIER'S POSITION TRACKING & HEALTH
CS860 MONITORING OPTIMIZATION MODEL USING
280 BIOSENSORS 120
281 CS863 SMART METERING GAUGE SYSTEM 120
SPEECH EMOTION RECOGNITION USING DEEP
CS865
282 LEARNING 121
EFFECTIVE APPROACHES OF E-COMMERCE
CS867
283 PRODUCT RECOMMENDATIONS 121
MOBILE CHARGING SYSTEM BASED ON COIN
CS868
284 DETECTION 122
NEW RECOMMENDER SYSTEM FOR ONLINE
CS869
285 COURSES USING KNOWLEDGE GRAPH MODELLING 122
MULTI-PLANT AND MULTI-CROP LEAF DISEASE
DETECTION AND CLASSIFICATION USING DEEP
CS870 NEURAL NETWORKS, MACHINE LEARNING, IMAGE
PROCESSING WITH PRECISION AGRICULTURE- A
286 REVIEW 122
A SURVEY ON DEEP LEARNING BASED VIDEO
CS880
287 SURVEILLANCE FRAMEWORK 123
PREDICTIVE ANALYTICS BASED MODELING OF THE
PURCHASE INTENTION OF ELECTRIC VEHICLES, AND
CS881 UNDERSTANDING THE DRIVERS AND RISKS IN
THEIR ADOPTION FOR THE PEOPLE OF TAMIL NADU
288 IN INDIA 123
289 CS887 BLOCKCHAIN BASED CROWDFUNDING 124
AI POWERED SMART IOT HELMET FOR THE EARLY
CS897
290 DETECTION OF COVID-19 AND ITS VARIANTS 124
291 CS899 VEHICULAR MISHAP ANALYSIS BASED ON SENSOR 125
STACKED BIDIRECTIONAL-LSTM NETWORK FOR
CS900
292 FAKENEWS DETECTION ON TWITTER DATA 125
MID-AIR GESTURE BASED MULTI-FINGER CONTROL
CS901 SYSTEM FOR PARALYZED PATIENTS USING LEAP
293 MOTION 125
INCORPORATING OF SECURITY METHODS INTO THE
CS904 SOFTWARE DEVELOPMENT LIFECYCLE PROCESS
294 (SDLC) 126
PREDICTION OF PHISHING CYBER ATTACKS USING
CS906
295 HYBRID DEEP LEARNING TECHNIQUES 126
METHODICAL TAMIL CHARACTER RECOGNITION
CS907
296 USING FABRICATED CNN MODEL 127
APPLICABLE TECHNIQUES FOR IMAGE
CS909
297 STEGANOGRAPHY: A SURVEY 127
CS913 DRIVER STATE MONITORING SYSTEM USING AI
298 127
PREDICTION OF LIVER DISEASE USING STATISTICAL
CS914
299 MACHINE LEARNING METHODS 128
SMART IRRIGATION SYSTEM FOR COIMBATORE
CS915
300 DISTRICT 128
MACHINE LEARNING BASED PLANT DISEASE
CS916 DETECTION USING CONVOLUTIONAL NEURAL
301 NETWORK 129
CROWDSOURCING FRAMEWORK FOR
CS917
302 AGRICULTURAL POWERED BY AI 129
THE COMPARATIVE STUDY OF CNN MODELS FOR
CS923 BREAST HISTOPATHOLOGICAL IMAGE
303 CLASSIFICATION 130
SOFTWARE DEFINED NETWORK FRAMEWORK &
CS927 ROUTING PROTOCOL BASED ON VANET
304 TECHNOLOGY 130
COMPARATIVE STUDY OF LINEAR PRECODING
CS934
305 TECHNIQUES 131
ELECTRONIC HEALTH RECORDS & DATA
CS936 MANAGEMENT USING HYPERLEDGER FABRIC IN
306 BLOCKCHAIN 131
OBTRUSION UNMASKING OF MACHINE LEARNING-
CS938 BASED ANALYSIS OF IMBALANCED NETWORK
307 TRAFFIC 132
HYBRID BILSTM NETWORK FOR IMPROVING CROP
CS939
308 PEST CLASSIFICATION 132
EMPIRICAL STUDY OF HUMAN FACIAL EMOTION
CS944
309 RECOGNITION: A DEEP LEARNING 132
A HYBRID MODEL APPROACH FOR STOCK MARKET
CS945
310 PRICE PREDICTION 133
MULTI-OBJECT DETECTION AND TRACKING USING
CS946
311 MACHINE LEARNING 133
AN IMPROVISED MACHINE LEARNING MODEL KNN
CS948
312 FOR MALWARE DETECTION AND CLASSIFICATION 134
FACTORS AFFECTING GAMING ENGINES:
CS951
313 ANALYTICAL STUDY 134
PREDICTING FLIGHT DELAY BASED ON
CS953
314 SENTIMENTAL ANALYSIS: MACHINE LEARNING 135
FORECASTING TEMPERATURE AND RAINFALL FOR
CS954 CROP CULTIVATION WITH CLASSIFICATION OF
315 GROUND WATER LEVEL 135
EMPIRICAL ANALYSIS ON CRIME PREDICTION
CS958
316 USING MACHINE LEARNING 136
TRAFFIC SIGN DETECTION USING HYBRID
CS962
317 NETWORK OF YOLO AND RESNET 136
HAND WRITTEN TEXT RECOGNITION AND GLOBAL
CS967
318 TRANSLATION 137
ATTEMPTING TO PERFORM ASSESSMENT FOR
CS971 PLANT BREEDING USING SCORED ENTROPY-BASED
319 CHOICE HIERARCHIE 137
ACCIDENT PREVENTION SYSTEM USING MACHINE
CS976
320 LEARNING 138
MONITORING AND ACCELERATING PLANT GROWTH
CS985
321 USING IOT AND HYDROPONICS 138
STOCK PRICE TREND ANALYSIS AND PREDICTION
CS987
322 OF CLOSING PRICE USING LSTM 139
REVIEW PAPER ON FACIAL RECOGNITION
CS994
323 TECHNIQUES 139
SECURE AND PROFICIENT PROVABLE DATA
CS1004 PROCURITY WITH PRIVACY PROTECTION IN CLOUD
324 STORAGE 139
IDENTIFICATION OF PNEUMONIA SYMPTOMS IN
CS1005 COVID19 PATIENTS USING TRANSFER DEEP
325 LEARNING APPROACH 140
VEHICLE LANE DETECTION PREVENTION AND
CS1007
326 SMART AUTODRIVE USING OPENCV 140
AN IOT FRAMEWORK BASED AUTOMATED
WIRELESS METER SYSTEM FOR
CS1008
MONITORING,BILLING AND CONTROLLING POWER
327 CONSUMPTION 140
THE ROLE OF ARTIOFICIAL INTELLIGENCE IN
CS1043
328 EDUCATION 141
DEEP LEARNING APPROACHES FOR ACCURATE
CS1044 SENTIMENTAL ANALYSIS OF ONLINE CONSUMER
329 FEEDBACK 141
EXPLORING THE INTERSECTION OF IOT AND
CS1045 BLOCKCHAIN:AN ANALYSIS OF SECURITY AND
330 PRIVACY RISKS 141
331 EC104 UNMANNED PLANT IRRIGATION SYSTEM USING IOT 142
DETECTING AND MITIGATING SELFISH PRIMARY
EC105
332 USERS IN COGNITIVE RADIO 142
IOT BASED DETECTION OF FOOD AND WATER
EC106
333 CONTAMINATION 143
OPTICALLY RECONFIGURABLE SLOTTED
EC1007 WAVEGUIDE ANTENNA ARRAY FOR 5G
334 APPLICATIONS 143
IMPROVING THE QUALITY AND PERFORMANCE OF
EC108 WIRELESS AD-HOC NETWORK BY USING MMCN
335 ARCHITECTURE 144
336 EC109 DEEP LEARNING BASED FAKE STAMP DETECTION 144
SPECTRUM SENSING FRAMEWORK AND ENERGY-
EC110 EFFICIENT RESOURCE ALLOCATION FOR COGNITION
337 ENHANCEMENT NETWORK 145
TRAFFIC ASSISTANCE AND CONTROL SYSTEM- AN
EC116
338 RFID APPROACH 145
DEVELOPMENT AND INTEGRATION OF DAPP WITH
EC123
339 BLOCKCHAIN SMART CONTRACT 146
DESIGN DEVELOPMENT AND EXECUTION OF SMART
EC124
340 CONTRACT 146
MEMORY EFFECTS IN DIGITAL FILTERS BASED ON
EC136
341 INPUT STRICT PASSIVITY 147
PEST DETECTION AND IDENTIFICATION ON PLANTS
EC149
342 USING CNN ALGORITHM: A SURVEY 147
INDUSTRY –SPECIFIC INTELLIGENT FIRE
EC151
343 MANAGEMENT SYSTEM: A SURVEY 148
DOES COVID PANDEMIC RAISE E-WASTE TO A
EC153
344 LARGER NUMBER? – ANALYSIS ANDPREDICTION 148
345 EC1006 DETECTION AND CONVERSION OF CO2 INTO MOIST 148
AN AUTONOMOUS CROP-CUTTING MECHANISM
EC1010
346 USING A DRONE 149
LEAN READINESS ASSESSMENT - TO INVESTIGATE
EC161 THE FEASIBILITY OF LEAN IMPLEMENTATION IN
347 SUGAR PRODUCTION PLANT 149
348 EC172 SURVEY ON SMART AGRICULTURE USING IOT 150
PERFORMANCE ANALYSIS OF TWO HORIZONTAL
EC201 FINS WITH STACKED GATE OXIDE OF TRIGATE
349 STRUCTURE 150
IMPLEMENTING HEALTH RECORDS USING
EC223
350 KNOWLEDGE GRAPHS AND BLOCKCHAIN 150
EDGE COMPUTATION ASSISTED GARBAGE
EC224 MONITORING AND ALERTING SYSTEM FOR A SMART
351 CITY 151
SUPPLY CHAIN MANAGEMENT IN MANUFACTURING
EC228
352 INDUSTRY USING INTERNET OF THINGS 151
SMART SEGREGATION OF WASTE AND AUTOMATIC
EC231
353 MONITORING SYSTEM 152
IOT-BASED SMART TRASH BIN FOR REAL-TIME
EC234
354 MONITORING AND MANAGEMENT OF SOLID WASTE 152
WASTE MANAGEMENT SYSTEM - A WASTAGE
EC235
355 CLASSIFIER USING DEEP LEARNING WITH IOT 152
BREAST TUMOR DETECTION USING DENSENET AND
EC246
356 KNN ALGORITHM 153
AUTOMATIC PRESSURE COOKER RISK DETECTION
EC263
357 AND CONTROL USING ARDUINO 153
358 EC264 IOT BASED SMART KITCHEN 154
A STRANGE MODEL FOR SECURE MEDICAL IMAGE
EC276
359 TRANSACTION - AN ATTRACTOR APPROACH 154
COLD SUPPLY CHAIN ADOPTING PI CONTROLLER TO
EC278
360 MAINTAIN TEMPERATURE DURING DRUG DELIVERY 154
REINFORCEMENT LEARNING ALGORITHM TO
EC279 SUSTAIN TIME IN PHARMA SUPPLY CHAIN
361 MANAGEMENT 155
ELECTRONIC VOTING MACHINE USING
EC287 FINGERPRINT SCANNER WITH IMAGE
362 ENHANCEMENT TECHNIQUE 155
DESIGN AND DEVELOPMENT OF UAV FOR
EC297 CHALLANGE FACED FOR DATA MULING PROBLEM
363 IN PATH 155
LP/CP RE-CONFIGURABLE ANTENNA FOR 5G
EC298
365 EC299 CUSTOM AND DESIGN OF AGRI DRONE 156
DUAL SEMI CIRCULAR DISC LOADED PLANAR
EC300
366 ANTENNA FOR 5G/WLAN-APPLICATIONS 156
MODIFIED ENCRYPTION STANDARD FOR
EC309 REVERSIBLE DATA HIDING USING AES AND LSB
367 STEGANOGRAPHY 157
A FLEXIBLE WIDEBAND SPIRAL ANTENNA USING
EC323 POLYDIMETHYLSILOXANE SUBSTRATE FOR
368 WEARABLE APPLICATIONS 157
IOT-BASED SMART INVENTORY MANAGEMENT
EC324
369 SYSTEM 158
DESIGN AND ANALYSIS OF ANNULAR SLOT
EC334 ANTENNA WITH MONOPOLE FOR WLAN USING
370 COMPUTER SIMULATION TECHNOLOGY 158
DESIGN OF DUAL BAND BODY AREA NETWORK
EC335
371 ANTENNA USING CST 158
DESIGN AND ANALYSIS OF M-SHAPED MONOPOLE
EC337
372 ANTENNA FOR BLUETOOTH COMMUNICATIONS 159
STEGANALYSIS USING CONVOLUTIONAL NEURAL
EC338
373 NETWORKS- YEDROUDJ NET 159
IMPACT OF MOBILITY AWARE RPL OBJECTIVE
EC352 FUNCTIONS IN MOBILE INTERNET OF THINGS
BLOCK LEVEL IMPLEMENTATION OF ORCA TOP
EC361
375 USING H -TECHNIQUE 160
IMPROVED RIEMANN SEQUENCE BASED SELECTIVE
EC371
376 MAPPING 160
IMPLEMENTATION OF BEAM SELECTION
EC378
377 ALGORITHM IN MILLIMETER WAVE MIMO SYSTEMS 161
NON-INVASIVE BLOOD GLUCOSE MONITORING
EC444
378 DEVICE USING ARDUINO 161
METAVERSE : CRYPTOCURRENCY PRICE ANALYSIS
EC480
379 USING MONTE CARLO SIMULATION 161
ANALYSIS OF MEMRISTOR BASED ANALOG
EC491
380 CIRCUITS FOR IMPROVED EFFICIENCY 162
A COMPARATIVE STUDY OF TRANSFER LEARNING
EC496 MODELS FOR OPTICAL SPECTRUM BASED OPTICAL
381 PERFORMANCE MONITORING 162
382 EC497 A IOT BASED SMART IRRIGATION SYSTEM 163
IOT BASED HEART RATE AND BLOOD OXYGEN
EC500
383 MONITOR WITH AUTOMATIC DATA SAVING 163
A DT FRAMEWORK BASED AI ROBOT FOR FACIAL
EC503
384 RECOGNITION USING IOT 163
PERFORMANCE ANALYSIS OF LOW-DENSITY
EC509
385 PARITY CHECK CODES FOR 5G TECHNOLOGY 164
IDENTIFICATION OF GENDER AND AGE USING
EC539 CLASSIFICATION AND CONVOLUTIONAL
386 NETWORKS 164
PERFORMANCE ANALYSIS AND COMPARISON OF
EC541 IMPROVED MFCC WITH HMM/VQ MODELING FOR
387 SPEECH RECOGNITION 165
SMART MCB DESIGN USING EMBEDDED AND IOT
EC566
388 SYSTEM 165
DEEP LEARNING FRAMEWORK FOR
EC574 IDENTIFICATION OF LEAF DISEASES IN NATIVE
389 PLANTS OF TAMIL NADU GEOGRAPHICAL REGION 165
DESIGN OF IOT AND MACHINE LEARNING BASED
EC581 MODEL FOR CROP PREDICTION AND FRUIT
390 RIPENESS DETECTION 166
CRITICAL REVIEW FOR ESTIMATION OF SNOW
EC592 DEPTH USING SATELLITE PRODUCTS AND
391 AUXILIARY PARAMETERS 167
101 SEQUENCE DETECTOR USING QCA
EC612
392 TECHNOLOGY 167
GRAPHENE BASED RECTANGULAR WIDEBAND
EC617 ANTENNA FOR HIGH SPEED AND SHORT DISTANCE
393 WIRELESS APPLICATIONS 167
A SURVEY ON SPECTRUM SENSING USING ENERGY
EC669
394 DETECTION FOR 4G WAVEFORM 168
REVIEW ON ANIMAL DETECTION FOR FARMLANDS
EC670
395 USING IMAGE PROCESSING AND IOT 168
A REVIEW: SECURE AND EFFICIENT DATA
EC671 AGGREGATION IN WIRELESS SENSOR NETWORKS
396 USING THE INTERNET OF THINGS 169
DESIGN OF METAMATERIAL INSPIRED ANTENNA
EC676 USING MACHINE LEARNING AND OPTIMIZATION
397 TECHNIQUES 169
INTELLIGENCE HARMLESS ANIMAL AVOIDER IN
EC677
398 AGRICULTURE FARM 170
FILTENNA WITH DEFECTED GROUND STRUCTURE
EC679
399 FOR 5G APPLICATION 170
BINARY BAT SWARM ALGORITHM FOR THINNING
EC682
400 ANTENNA ELEMENTS IN LINEAR ARRAY 170
INTERFACING KERA’S DEEP LEARNING TECHNIQUE
EC683
401 FOR REAL-TIME AGE AND GENDER PREDICTION 171
PARKINSON’S DISEASE CLASSIFICATION USING
EC693
AUTOMATIC SHOPPING CART USING RFID AND
EC698
403 NODEMCU 171
A FORK SHAPED UWB ANTENNA WITH
EC700 WLAN/WIMAX/ X BAND REJECTION
404 CHARACTERISTICS 172
CLASSIFICATION AND DETECTION OF SKIN CANCER
EC701
405 USING DEEP LEARNING METHODS 172
DESIGN AND IMPLEMENTATION OF FUZZY LOGIC
EC706 CONTROLLER FOR BOILER TEMPERATURE CONTROL
406 USING LABVIEW 173
DESIGN AND ANALYZING OF HEXAGON-SHAPED
EC718 MICROSTRIP PATCH ANTENNA FOR BIOMEDICAL
RECONFIGURABLE HARDWARE IMPLEMENTATION
EC719 OF CNN ARCHITECTURE USING ZEROBYPASS
408 MULTIPLIER 174
A TRAFFIC ROAD SIGN RECOGNITION SYSTEM
EC720 USING RASPBERRY PI WITH OPEN CV LIBRARIES
409 BASED ON EMBEDDED TECHNOLOGIES 174
REVIEW ON CARDIOVASCULAR DISEASE
EC721 PREDICTION USING MACHINE LEARNING
410 ALGORITHM 175
EFFICIENTLY ENHANCING ENERGY BY MEANS OF
EC723 LEACH PROTOCOL FOR A WIRELESS SENSOR
411 NETWORKS 175
SMART SOLID WASTE COLLECTION FOR A
EC724
412 SUSTAINABLE SMART CITY USINGARDUINO UNO 175
A NOVEL FRAMEWORK ON QOS IN IOT
EC725 APPLICATIONS FOR IMPROVISING ADAPTABILITY
413 AND DISTRIBUTIVENESS 176
A NOVEL INTERFERENCE MITIGATION TECHNIQUE
EC727 AND SIGNAL STRENGTH PREDICTION IN FUTURE
414 WIRELESS COMMUNICATION 176
BLOCK CHAIN BASED IDENTITY AUTHENTICATION
EC728
415 SYSTEM FOR A WIRELESS SENSOR NETWORKS 177
DIGITAL FILTER DESIGN ON HIGH SPEED
EC729
416 COMMUNICATION WITH LOW POWER CRITERIA 177
SYSTEM FOR REAL-TIME MONITORING AND
EC735 REPORTING OF THE WEATHER BASED ON THE
417 INTERNET OF THINGS 177
418 EC740 COAL MINERS ACTIVITY SENSING USING IOT 178
DEVELOPMENT OF IOT BASED SOLAR WATER
EC741
419 HEATER CONTROLLER 178
420 EC742 VIRTUAL MOUSE 179
IMAGE REFLECTION REMOVAL USING ITERATIVE
EC758 BOOST CONVOLUTIONAL LSTM NETWORK
421 THROUGH FEATURE LOSS 179
DESIGN OF T AND Y MICRO-CHANNEL THROUGH
EC776
422 MULTI PHYSICS COSMOL 179
423 EC777 EMOTION ANALYSIS OF TWEETS 180
OPTICAL IMAGING FOR TELEPHOTO APPLICATIONS
EC778 FEATURING A LENS DESIGN OPTIMIZED FOR
424 THERMAL COMPENSATION 180
RECONFIGURABLE MIMO ANTENNA FOR 5G
EC782
RFID-ENABLED WIRELESS PRINTING SOLUTION
EC783
426 USING RASPBERRY PI 181
DEEP LEARNING BASED INTELLIGENT ALERT
EC784
427 SYSTEM FOR VISUALLY IMPAIRED PEOPLE 181
QUADCOPTER NAVIGATION USING TELEMETRY
EC785
428 DATA 182
DATA ANALYTICS TO REFLECT THE CURRENT
EC787
429 POLITICS ON TWITTER USING MACHINE LEARNING 182
A MEDICAL ASSISSTING SYSTEM USING AUGURED
EC791
430 REALITY FOR CRITICAL ANALYSIS 182
ANALYTICAL REVIEW OF DEEP NEURAL
EC793
431 TECHNIQUES FOR BREAST TUMOR DETECTION. 183
REGULATING AND MONITORING IOT CONTROLLED
EC796
432 SOLAR POWER PLANT BY 183
WIRELESS MEDICAL SENSOR NETWORK BASED
EC808 HEALTHCARE MONITORING SYSTEM IN NARROW
433 BAND IOT 184
434 EC809 SMART LIBRARY USING ARDUINO 184
DESIGN AND IMPLEMENTATION OF IOT-BASED
EC843 PROGRAMMABLE ASSISTIVE INTELLIGENT
435 ADAPTIVE PILLBOX 184
ENERGY EFFICIENT SMART STREET LIGHT SYSTEM
EC851
436 BASED ON PULSE WIDTH MODULATION 185
IOT-BASED INTELLIGENT MODELING OF SECURITY
EC854
437 MONITORING SYSTEM FOR HUMAN SAFETY 185
ARTIFICIAL INTELLIGENCE BASED RECOGNITION OF
EC862
438 LIVER MELANOMA USING SONOGRAM IMAGES 186
DESIGN AND OPTIMIZATION OF MICROCHANNEL
EC871
439 FOR NEONATAL 186
DETECTION AND SEGMENTATION OF RETINOPATHY
EC874
440 DISEASES USING EAD-NET WITH FUNDUS IMAGES 187
A REVIEW ON INFECTIOUS AND CARDIOVASCULAR
EC876
441 DISEASE CLASSIFICATION USING DEEP LEARNING 187
NOVEL, CLOCK GATING BROADCASTING
EC893 APPLICATIONS FOR LOW-POWER FPGA
442 ARCHITECTURES 187
HEALING AND PREVENTING TREES FROM BEETLES
EC902
443 USINGPESTICIDES 188
VEHICLE DETECTION AND CLASSIFICATION USING
EC908
444 DEEP LEARNING FOR HAZY IMAGES 188
A HIERARCHICAL WASTE SEGREGATION USING
EC924
445 SMART DUSTBIN 189
ARTIFICIAL INTELLIGENCE FOR IOT-BASED
EC943
446 HEALTHCARE SYSTEM 189
AGE GROUP DETERMINATION USING LOGISTIC
EC975
447 REGRESSION THROUGH A BIOMETRIC APPROACH 189
ADVANCE IOT BASED AIR POLLUTION DETECTION
EC988
448 AND CONTROL SYSTEM IN VEHICLES 190
449 EC989 SENSOR BASED SEWAGE SEGMENTATION 190
A SYSTEMATIC TECHNIQUE USED FOR PORTABLE
EC990
450 TELEMEDICINE SYSTEM 190
COMPARATIVE STUDY TO DETERMINE ACCURACY
EC991 FOR THE WEATHER PREDICTION USING MACHINE
451 LEARNING 191
SMART USE OF SOLAR RADIATION IN AGRICULTURE
EC992
452 PURPOSE 191
A REVIEW ON THE ANALYSIS OF 5G TECHNOLOGY
EE171
453 AND ITS IMPACT ON HUMANS 191
DESIGN OF MULTILEVEL INVERTOR WITH ITS
EE240 STUDY OF DIFFERENT CONFIGURATIONS AND
DESIGN OF ENERGY EFFICIENT LOW VOLTAGE
EE486
455 SINGLE PHASE IM 192
GENETIC BASED FUZZY LOGIC CONTROL OF SEDC
EE487
456 MOTOR 192
IMPLEMENTATION AND CLASSIFICATION OF
EE553 BREAST CANCER HISTOPATHOLOGICAL IMAGE
457 USING SUPPORT VECTOR MACHINE LEARNING 193
MODIFIED PI CONTROLLER FOR BLDC MOTOR IN
EE578
458 ELECTRIC VEHICLE APPLICATIONS 193
AN OVERVIEW ON DEVELOPMENT AND PROSPECTS
EE656
459 OF GEOTHERMAL POWER PLANT IN INDIA 193
REVIEW OF HYBRID WIND SOLAR PV TECHNOLOGY
EE673
460 IN THE GENERATION OF ELECTRICITY 194
461 EE674 PORTABLE WATER QUALITY DETECTOR 195
SOLAR PV FED MODULAR MULTILEVEL INVERTER
EE921 WITH FUZZY LOGIC CONTROL FOR MARINE WATER
462 PUMPING APPLICATIONS 195
A NEW FAULT TOLERANT FIVE – LEVEL
EE973
463 MULTILEVEL INVERTER FOR CRITICAL LOADS 195
PERFORMANCE INVESTIGATION OF RESONANT
EE977
464 CONVERTER WITH AND WITHOUT CONTROLLER 195
CONSTANCY INVESTIGATION OF FUZZY
EE978 CONTROLLER OVER PROPORTIONAL INTEGRAL
465 CONTROLLER FOR RESONANT CONVERTER 195
ISBN: 979-8-3503-4821-7
PAPER : CS100
Au A Blockchain Based Certificate Authentication System

G.Balamurugan, K. Kishore Anthuvan Sahayaraj
[email protected]
ABSTRACT— The certificates of the Secondary School Leaving Certificate (SSLC), the Higher
Secondary Leaving Certificate (HSLC), and academic certificates are all digitized at the academic
institution and made available to the students in today's digital environment. The rise in the number of
security breaches creates a threat to the users' right to privacy regarding their academic digital certificates.
Validation and verification of the digital certificates is very challenging for the institution and the
business. The system can able to give a digital certificate validation that is both safer and more efficient
thanks to the use of blockchain technology. The purpose of this proposed system is to provide a suggestion
for a certificate administration and verification system that, by using blockchain technology, has the
ability to provide a viable answer for the problem of academic credential issuance and verification. Hash,
public-private key cryptography, mining peer-to-peer networks, and proof of work are only few of the
capabilities that are included in blockchain technology.
Keywords-: Blockchain, Digital Certificate, Certificate Verification, Hyper ledger, Hashing.
PAPER : CS115
Hand Gestures for Mouse Movements with Hand and Computer Interaction model
Kumaresan R, Dr. Kalaivaani P.C.D, Manow Ranjith A J, Nandha Kumar S
[email protected]
ABSTRACT: The design and implementation of interactive computing systems that humans can engage
with is known as human-computer interaction (HCI). It encompasses both desktop and embedded systems,
found in a variety of devices. Humans interact with computers through input devices such as the touch
pad, mouse, joystick, and keyboard. Despite the fact that input devices have improved over the years,
many people still find interacting with computers and gadgets to be a frustrating experience. Intelligent
gadgets that work alongside computers are being developed now, making Human-Computer Interaction
simpler. Keyboards may be challenging for disabled persons to utilise in ordinary chores. Diagramming
with a keyboard is not possible. The proposed work, attempts to develop a gesture-based system that
allows users to control desktop mouse actions. It recognises hand gesture motions using a desktop
webcam. This utilises a primitive camera to do mouse cursor activities instead of standard mouse device.
The Virtual Mouse is a camera-based infrastructure that connects the user to the system. It allows users
to control mouse motions and connect with machines without the usage of physical devices. By
minimizing human contact with public devices such as elevators and ATMs, reduces the spread of
contagious diseases in the present scenario.
Keyword: opencv, threshold, mediapipe, gesture recognition, autopy, virtual mouse.
PAPER : CS117
ARTIFICIAL INTELLIGENCE DEPLOYMENT AND THE POTENCY OF HRM FUNCTIONS
Dr. A.P.Gayatri, Dr. P. Lavanya, Mrs. Ch.Rajitha, Naveen Kumar Butta, Dr. U.J. Supraveen
ABSTRACT—The overall objective of the current study is to investigate the causality between the
application of AI and the efficiency of human resource management (HRM) activities. The HRM
1
ISBN: 979-8-3503-4821-7
activities covered by this research are talent acquisition, HR analytics, and recruitment and selection. This
study employed a quantitative research design. Managers and staff working for companies in India's three
main cities, New Delhi, Mumbai, and Kolkata, completed a questionnaire with 30 questions. Data were
gathered through the use of random sampling. The 150 participants' data were analyzed by the study using
SPSS software. This study's objective is to investigate the causality between the application of AI and the
efficiency of HRM activities. The HRM activities covered by this research are talent acquisition, HR
analytics, and recruitment and selection. This study employed a quantitative research design. Data were
gathered through the use of random sampling. The 150 participants' data were analyzed by the study using
SPSS software. The outcome of the analysis depicts that there exist a strong correlation between few
HRM functions and AI implementation.
Keywords: Functions of HRM; AI; HR Analytics; Talent Acquisition; Recruitment and Selection; SPSS;
PAPER : CS120
The Approach of Digital Marketing is Being Revolutionized by Machine Learning Possibilities
Dr. K.Jagannayaki, Dr. Mantha Srinivas, Dr. K.Sharath Babu, Naveen Kumar Butta, Dr. Sayyad
Saadiq Ali,
ABSTRACT— As a result of changing customer behaviour, the proliferation of data, and the ongoing
evolution of DM, marketers are sometimes forced to wade through massive amounts of data that might not
even provide them with the overall picture they really ought to make impactful business choices. Eighty-
four percent of marketing businesses are introducing or increasing their usage of ML in 2018[1] following
the transformation of ML algorithms in other real-world applications. It becomes simpler to accurately
forecast and evaluate consumer behaviour. In our work, we first build a state of the art on the primary and
most prevalent ML potentials in DM strategies, and we then demonstrate how ML technologies may be
employed at a big scale for marketing by analysing extraordinarily huge data sets. They are able to better
understand their target customers and maximise their contact with them because to the way that ML is
incorporated into digital marketing methods.
Keywords— Digital Marketing; Machine Learning; Customer Behaviour; Predictive Analysis;
PAPER: CS126
New Real Time Clinical Decision System Using Machine Learning – Disease Prediction
K.Pushpa Rani, Ch.Sabitha, P.Sreedevi, A.Sangeetha
[email protected]
Abstract— Deep learning can be known as an area of machine learning which is useful to reorganize
patient care. It is not yet part of standard care, especially when it comes to individual patient care. It is not
clear to what extent various data-driven techniques are being used to support clinical decision making
systems (CDS). Therefore, there has not been a review of ways in which research in machine learning and
other types of data-driven techniques can contribute effectively to clinical care. There are many different
parts in clinical decision making. Few of them are disease predictions, online patient tracking; artifact
detection, state estimation etc. There are many diseases present now-a-days. Every disease has its own
symptoms. Our project is to find the disease based on the symptoms. The disease prediction is made using
machine learning techniques.
Keywords- Clinical, decision, system, patient, diseases
2
ISBN: 979-8-3503-4821-7
PAPER: CS128
Diagnosis of Lung and Pleural Diseases by Machine Learning Algorithms
B.Devananda Rao, Dr.Mahammad Arshad
[email protected]
Abstract— This present work employs the classification equipment for detection of pleural carcinoma with
an accuracy of 98.30%. The data collected from the lung cancer patients is divided into health data and
clinical data for analyzing the efficacy of the Algorithms of Machine learning. To direct the physicians in
early detection and providing suitable treatment for the lung cancer patients, the machine learning
algorithms are utilized and analyzed based on the data collected from the health and medical data of the
patients. Separation process applied with various algorithms of machine learning to increase the success
levels. To attain about 98.30% of success rate many algorithms are evaluated. Among the numerous
algorithms tried the most productive isolation process is the Linear Discriminant Analysis.
Keywords- Pleural carcinoma, lung cancer detection, classification, machine learning algorithms, linear
discriminant analysis.
PAPER: CS133
Recognizing Handwritten Offline Tamil Character using VAE-GAN & CNN
N sasipriyaa, dr.p.natesan, dr.r.s.mohana, arvindkumar p, arwin prakadis r s, aswinsurya k
[email protected]
ABSTRACT--Handwritten character recognition (HCR) is a pattern recognition problem. The HCR
complication has been extensively learned for several decades, but there has been tiny research on
independent script models. This is due to a variety of variables, lack of datasets, the concentration of most
traditional research on character recognition approaches are language specific and
unattainable. Convolutional Neural Networks are deep learning models that take input and process it
through multiple layers to produce the desired output based on training. In order to increase the volume of
the dataset, GAN (Generative Adversarial Network) framework was adapted on the HP Labs e TAMIL
handwritten character dataset. Despite the fact that GAN improves the dataset, processing the images takes
a long time, the proposed work hybrid the models Variational Autoencoder (VAE) and GAN model to
speed up processing. A GAN tries to produce new data that cannot be identified from real data, whereas a
VAE learns to encode the provided data and then reconstructs the new images from the encoding. IN VAE-
GAN, we employ the latent representations produced by an encoder for a variety of purposes. The proposed
system uses VAE-GAN to enlarge the dataset and Convolutional Neural Networks to recognise handwritten
Tamil characters. The accuracy of VAE-GAN with CNN is found to be 86%.
Keywords: Deep Learning, VAE GAN, Tamil, CNN, character recognition
PAPER : CS134
Blockchain Applications in Healthcare- Recent Advancements
Kanika Agrawal, Mayank Aggarwal
[email protected]
ABSTRACT— In recent years, new inventions are taking place to adapt blockchain technology in various
domains. Different researchers are proposing different frameworks for many appli-cations. From the
research done so far, it was seen that there was the need to focus on in-depth blockchain application analysis
3
ISBN: 979-8-3503-4821-7
in healthcare sector. Motivated from the aforementioned facts, this research paper delivers blockchain’s
effectiveness and usefulness in the emerging domain of healthcare. The authors have focused on the
researches done in last three years and identified a number of applications of blockchain technology in
healthcare, a few of them discussed are – biomedical, prediction of disease, electronic health records,
genomics and pharmaceuticals. Additionally, various challenges faced by the traditional healthcare sector
along with the blockchain possible solutions are also discussed. Hence, this paper helps researchers to
provide insights about works done in blockchain and its application in healthcare sector.
Keywords- genome, healthcare, health records, pharmaceu-ticals, supplychain
.
PAPER : CS135
Blockchain based Healthcare System to Secure Health Records
Kanika Agrawal, Mayank Aggarwal
[email protected]
ABSTRACT— With the advancement in technology, blockchain is playing a major role all around the
world. The blockchain applications shows it’s significance in many applications such as financial services,
aircraft industry, e-voting mechanisms etc. However, the healthcare sector is focusing more attention
towards the technology to utilize the unique features of blockchain like security, decentralization and trust.
Motivated from the aforementioned facts, this paper proposes MyEasyHeathcare, an efficient healthcare
model to help provide the patients a secure facility from the hospital. The model helps doctor and patient
to register and then authorizes a particular doctor to a particular patient. The physician or pathologist checks
the patients accordingly and gives prescription as treatment. This paper also discusses research challenges
faced by traditional systems and their possible blockchain solutions.
Keywords- Blockchain, Healthcare, Respiration Rate, Serum Calcium, Treatment.
PAPER : CS150
A Systematic Investigation of Cyber Harassment Intention Behaviors and Its Impacts on
Social Media Platforms
S. Abarna, J.I. Sheeba , S. Pradeep Devaneyan
[email protected]
ABSTRACT— Several technologies have been able to fundamentally affect our lives in just a few decades
by integrating themselves into our society. News and opinion are now mostly shared on social media. There
are both positive and negative aspects to this. Nowadays, practically everyone has a smartphone or other
internet-accessing gadget. Modern technology becomes available to kids at progressively younger ages.
The majority of them use social media, which contributes to the rise of cyber harassment and the negative
aspects of the internet. Due to the ease of access to these platforms, cyberbullying is prevalent on social
media. Moreover, when dealing with others, the perpetrators have the ability to remain anonymous.
Cyberbullying statistics for 2022 highlight the alarming incidence of harassment in the virtual world, and
they are disturbing. According to these data, Cyberbullying sufferers are more than twice as likely to self-
harm or entertain suicidal thoughts as a coping mechanism. Anxiety and sadness might develop in certain
victims of cyberbullying. Psychologists believe that social media has contributed to an increase in stress
and depression over the last decade. On the other side, the most typical targets for online bullies are young
children, women, and individuals with non-traditional sexual orientations. Students are often unable to
concentrate on their studies because they are afraid of the trolls. As a result, this Research Essential focuses
on Cyber harassment statistics and facts in order to raise awareness among parents, students, and authorities
4
ISBN: 979-8-3503-4821-7
worldwide and reveal alarming facts about virtual harassment, the intention of the harasser, its impact, the
relationship between victim and harasser, and its many different shapes and forms.
Keywords- Cyber harassment, bullying, Intention Behavior, Impacts, Digital Platform.
PAPER : CS152
Stock Price Prediction using Random Forest Classifier and Backtesting
Venkata Sai P Bhamidipati, Mrs. D. Saisanthiya
[email protected]
ABSTRACT Stock price prediction is an attempt to determine the future price of company stock and(or)
other categories traded on the National Stock Exchange[NSE]. A stock's future price forecast that is
accurate might result in a sizable profit. The prediction of the price of a stock is a very important topic
considering that the raw data is ambiguous. Though many consider this to be an ideal machine learning
problem, the solution for it is far from perfect. Many people are actively researching/funding research to
derive a solution for this problem.
Usually the methodologies for prediction fall into three broad categories which can (and might) overlap:
1. Core research (Manual evaluation): Core research (or) Manual evaluation is a procedure by which
investors of the stock market manually calculate the returns on their invested amount by estimating the
stock price based on news articles, market movement, etc.
2. Technical Research (Tracking): Technical Research (or) Tracking is a procedure in which the stock is
monitored using a software application which can track the stock value according to historical data and
displays relevant information for investors.
3. Technical Techniques (Machine Learning): Technical Techniques (or) Machine Learning is a procedure
by which an application is enabled to become more accurate at predicting outcomes. With the use of these
techniques, investors can get a better-estimated prediction.
We aim to use machine learning as a procedure and overlap it with a concept of tracking known as
backtesting to provide stock price predictions.
Keywords- Stock price prediction; Machine learning; Backtest
PAPER : CS155
Comparison of Cryptography algorithms in Blockchain
Amu.D, Dr.Santhi Baskaran
[email protected]
ABSTRACT— Various algorithms have been employed to enhance the performance of blockchain.
Successful algorithms have been developed with the potential to use blockchain technology. In this study,
we analyze and explain a list of significant blockchain-based algorithms, including Rivest Shamir
Adleman, Advanced Encryption Standard, Data Encryption Standard, Digital Signature, and Elliptic Curve
Cryptography. In this article, we examine the Quantum Teleportation security algorithm for Blockchain
Technology. This technique is based on encryption to provide well security.Quantum teleportation
compares favorably to RSA, DES, AES, and ECC. This survey document serves as a valuable guide for
upcoming research issues in Blockchain Technology.
Keywords— Blockchain, Encryption, Cryptography algorithm
5
ISBN: 979-8-3503-4821-7
PAPER : CS162
BITCOIN PRICE PREDICTION USING MACHINE LEARNING
Dr.S.Nithiya, G.Parimala
[email protected]
ABSTRACT— Bitcoin is a type of Internet currency that is both a digital asset and a payment method. It
enables anonymous payment from one person to another, making it a popular payment mechanism for
online illegal activity. Due to its recent price increase, Bitcoin has attracted a lot of attention from the media
and the public. The purpose of this study is to discover the Bitcoin price’s predictable price direction.
Machine learning models are likely to provide us with the information we require to Understand the future
of cryptocurrency. It won't tell us what will happen in the future, but it might show us the overall trend and
direction in which prices are likely to move. The proposed Methodology aims to create a machine learning
model that uses data to learn about the patterns in the dataset and then uses a machine-learning algorithm
to forecast the bitcoin price.
PAPER : CS163
HUMAN ACTIVITY RECOGNITION FOR ANALYSING FITNESS DATASET USING A
FITNESS TRACKER.
Soumiya Chadha, Ishita Raj , D. Saisanthiya
[email protected]
ABSTRACT—The rise in eminence in the research based on Human Activity Recognition (HAR) due to
its extensive scope in its application such as monitoring activities of daily life, fitness, security, and many
more areas, is no mystery. In this project, we will be concentrating on the application of HAR in mobile
health. The demand for HAR has increased in recent times due to the rapidly growing elderly, obese, and
overall unhealthy population. We will be dealing with a Smartphone-based HAR system as a fitness tracker
that uses high-dimensional sensors to collect fitness datasets and infer a confusion matrix for HAR in
different activities to analyze the best fit in this work and predict the most accurate activity. It is crucial to
select the appropriate features to support our research. We have used the Random Forest feature selection
and the extracted data was analyzed using different classification techniques along with the previously
mentioned scheme to generate the most accurate results and find a method that can be used by developers
in the fitness tracker industry. Our data is a collection of 30 participants within an age bracket of 19 – 48
years performing six activities namely, Walking, Walking Upstairs, Walking Downstairs, Sitting, Standing,
and Laying while carrying a waist mounted smartphone with embedded wireless sensors. These activities
recorded certain movements, acceleration, velocity, and heart rate which was compiled into readable data
and worked upon.
Keywords—Analysis, Smartphone, Random Forest, Machine Learning, Classification, Sensors
PAPER : CS164
Analysis, Smartphone, Random Forest, Machine Learning, Classification, Sensors
M. Rajalakshmi, Akshat Khare, Pranjal Patel
[email protected]
ABSTRACT— The idea behind this research idea is to develop a deep learning model for traffic vehicle
detection, which predicts types of vehicles (Truck, Car, Bike) using frame difference through OpenCV
library. The aim of this research work is to help CCTV surveillance capture the weight of in and out traffic
in highways(national and express), which can be used by engineers to analyze the traffic. Many feature
extraction techniques has been applied such as Frame Difference, Contour mapping to draw out insightful
features for model to predict on.
6
ISBN: 979-8-3503-4821-7
PAPER : CS167
An Effective Sentiment Analysis Classification for Opinion Prediction using Machine Learning
Schemes
Dr.T.M.Saravanan, Dr.M.Jagadeesan, Dr.P.A.Selvaraj, B.M.Rithika, G.Karan, R.C.Tamil Selvan
ABSTRACT—In the blooming of web technology, the rise of electronic commerce websites is rapidly
increasing. Users started expressing their views similar to text reviews or comments. This research deals
with the identification and extraction of the exact opinion of the customers across reviews, feedbacks, or
comments. Pre-processing techniques are used to clean the dataset and the machine learning algorithm like
k-nearest neighbor (KNN) is used as a solution to this problem, to have a good evaluation of sentiments
(Positive, Negative, Neutral). In accumulation, the precision of the algorithm highly depends on the value
of k and distance metric for classification. In our work we used k value of 6 and euclidean distance equation
for classification of the KNN algorithm. The proposed KNN algorithm has been applied on publicly
available standard review dataset repository, to show the exact nature and features based on customer’s
opinions, and the accuracy of determining the feasibility of the k-nearest neighbor algorithm of 98% is
predicted.
Keywords-Sentiment Analysis, KNN, Euclidean Distance, Machine Learning and Opinion.
PAPER : CS175
Sentiment Analysis, KNN, Euclidean Distance, Machine Learning and Opinion.
Ashutosh Shukla, Amrit Aanand, Dr. S.Nithiya
[email protected]
ABSTRACT—A significant problem in machine learning is automatic voice recognition, particularly
continuous speech recognition with a vast vocabulary. The standard speech recognition framework has
been the hidden Markov model (HMM)-Gaussian mixed model (GMM) for a long time. However, more
recently, the performance of the HMM-deep neural network (DNN) model and the end-to-end model
employing deep learning has surpassed that of the HMM- GMM.
Keywords—LibriSpeech, CNN, HMM, Acoustic and Language Modeling
PAPER : CS189
Comparative Study of Detection and Analysis of Different Malware with the Help of Different
Algorithm
Dr.P.A.Selvaraj, Dr.M.Jagadeesan, Dr.T.M.Saravanan,Aniket Kumar, Anshu Kumar, Mayank
Kumar Singh
[email protected]
ABSTRACT- The motive of this paper is to detect the malware from computer systems in order to protect
the confidential data, information, documents etc. from being accessing. The detection of malware is
necessary because it steals the data from that system which is affected by malware. There are different
malware detection techniques (cloud-based, signature-based, Iot-based, heuristic based etc.) and different
malware detection tools (static, dynamic) area used in this paper to detect new generation malware. It is
necessary to detect malware because the attacks of malware badly affect our economy and no one sector is
untouched by it. The detection of malware is compulsory because it exploits goal devices vulnerabilities,
along with a Trojan horse in valid software e.g. browser that may be hijacked. There are also different tools
used for detection of malware like static or dynamic that we see in this paper. We also see different methods
of detection of malware in android.
7
ISBN: 979-8-3503-4821-7
PAPER : CS197
Chatbot based Virtual Diagnosis Assistant for Skin Diseases and Bone Fracture Detection using
Neural network
[email protected]
ABSTRACT— The motto of developing the diagnosis assistant is to save time and money in almost
possible situations. Nowadays people using the internet has become an hourly habit due to which their busy
schedule leads them to avoid visiting the hospital so they can use a chatbot to get their medical queries. It’s
a Rule based chatbot develop with assistance of network communication to diagnose skin and ortho
diseases. The information is trained in the neural network-based system. The trained information helps
doctors to identify the disease accurately.
Keywords - Image Classification, Convolution Neural net , TCP/IP Network , Rule based NLP
PAPER : CS199
Scaling Blockchain Systems using DAG
Parth Dhawan, Arjav Rastogi , D. Sai Santhiya
[email protected]
ABSTRACT— Directed acyclic graphs are used for one primary reason: they make it easier to build
several blocks at once. The number of forks in a blockchain naturally grows as the rate at which blocks are
generated increases. As a consequence of this observation, several experts have come to the opinion that
organising the resulting directed acyclic network is the most effective way to solve the scalability issue that
plagues Bitcoin and other similar cryptocurrencies. A significant difficulty for this system arises from the
fact that blocks created concurrently or in close proximity may contain similar or conflicting transactions,
all of which must be handled and organised in linear order. The idea of using several blockchains, each of
which processes transactions independently based on the Sender's address, is presented as a possible
alternative method. Consensus algorithms continue to be the limiting factor in the size of a blockchain
network and the speed with which data can be published, despite the advancements in blockchain
technology and the advent of new use cases. Despite blockchain's ability to accommodate a greater variety
of data, this remains the case. Immediate action is required to address the critical problem of how to boost
alliance blockchain performance in a secure and efficient manner.
Keywords— Trust based-Practical Byzantine Fault Tolerance(tPBFT); Blockchain Technologies(BCT);
Proof of Work(PoW); Directed Acyclic Graph(DAG); Bitcoin; Blockchain; Cryptocurrencies; Transaction
Per Second(tps); Time To Finality(TTF); Nakamoto Consensus; Forking; Side Chains; Main Chain.)
PAPER : CS202
Lane Coloration in Real Time Vehicular Ad-hoc Networks (VANETS)
Arjit Vatsa, Sanskar Agarwal, Ranjani M
[email protected]
ABSTRACT- With the advent of emerging technologies, it is important for every aspect of living to
become fast – paced and safer. Working in correspondence to above idea, we are creating a system of Road
Lane coloration and detection suitable for Indian Roads. For this we will be applying various ideas such as
gamma – correction (for darker roads), addition of Kalman Filter, Canny Edge Detector (for edge detection)
and DBSCAN (to cluster the detected lines). AI aspects have made it possible to build self-sufficient
vehicles and GPS, but if there exists a system by which we can make this lane detection and self-driving
car scenarios safer, wouldn’t it be better? Since our goal is to provide self-driving technology via a mobile
device, we were looking for a solution that would be less computationally intensive and operate in real time
from a phone. It was difficult for us to recognize road lanes during the day, so we started doing it at night.
8
ISBN: 979-8-3503-4821-7
We are trying to solve this issue with the addition algorithms and by removal of as much noise. For our
system, White Gaussian noise is used. We have estimated the state's error using this noise model, and then
we utilized the estimator to get the anticipated state using the measurement noise. The Kalman filter
averages out variations in lines that are detected along the lanes by summing up measurement error and
prior state. Since identified lane marker lines have a predictive feature from a prior state, they remain stable
over time. Additionally, even under extremely low light conditions, it is still able to detect lanes by recalling
previously recognized lanes from a previous video frame. We have begun to work on multi-lane detection,
tracking, approximating road lane width, etc. You could soon be able to use your phone to access a high-
end perception system.
Key words: Canny Edge Detection, Kalman Filter, Lane Coloration, Deep Learning, Indian Roads,
DBSCAN
PAPER : CS221
Canny Edge Detection, Kalman Filter, Lane Coloration, Deep Learning, Indian Roads, DBSCAN
Roopa Devi E.M , R.Shanthakumari, R.Rajadevi, Anoj Roshan M, Hari V, Lakshmanan
[email protected]
ABSTRACT-- Electronic Health Records have enabled the biosciences to advance more significantly and
produce vast amounts of data. Due to this massive amount of data, there is an essential necessity for
knowledge development. In this area of biosciences, machine learning has a significant impact. The
inability of the kidney to carry out its normal functions, such as excretion and regulating the body's
electrolyte balance, is known as Chronic Kidney Disease (CKD), otherwise called chronic kidney failure.
CKD may be brought on by an ancestral record of renal disease, excessive blood pressure or diabetes. The
likelihood of this kidney dysfunction deteriorating further over time is substantial. Heart conditions,
anaemia, bone conditions, excessive potassium and calcium levels, and anaemia are among the extremely
frequent consequences that arise with kidney failure. To predict CKD, a solid prediction algorithm is
needed. This study predicts CKD using data pre-processing, Particle Swarm Optimization (PSO) based
feature selection and a variety of classifiers including ensemble classifiers applied on the chronic kidney
disease dataset obtained from the UCI Repository. It provides an optimum prediction model for detection
of CKD. The outcomes of the framework indicate positive signs of improved CKD prediction.
KeyWords— Chronic Kidney Disease, Particle Swarm Optimisation, Decision Tree, Random Forest,
XGBoost, AdaBoost.
PAPER : CS226
Implementation of Yara rules in Android
Pragya Bharti, Shreya Saha Roy, Dr. A. Suresh.
[email protected]
ABSTRACT— Malwares are malicious softwares aimed to damage and destroy computer systems and
networks. Malware can exist in a wide variety of devices and operating systems. Cryptographic hashing
and fuzzy hashing are two types of signature-based malware detection and classification techniques. In this
paper we have tried to study the implementation of YARA rules in Android operating system, the properties
of YARA rules and how it helps in an efficient detection of malicious android applications in the market.
We outline the syntactical structure of YARA rules, their use cases, and how to create a YARA rule for a
single malware or a family of malwares using Androguard and Cuckoo.
Keywords— Malware, operating systems, signature-based, cryptographic hashing, fuzzy hashing, YARA,
androguard, cuckoo.
9
ISBN: 979-8-3503-4821-7
PAPER : CS227
An Ensemble of Machine and Deep Learning Models for Real Time Credit Card Scam Recognition
Khushi Goyal, Shaurya Singh, Muskan Gulati, Dr. A. Suresh
[email protected]
ABSTRACT— As the E-commerce sector is getting large, the use of electronic money and is getting wider
and wider. Credit cards are the most useful and easy tools for payment. It is easy to use and reduces the
efforts made by humans. But with advantages some disadvantages also come hand in hand. Many frauds
take place while making the transactions and due to this many people lose millions of money. Hence, there
need to be a detection system so that people can make the transactions without the fear of frauds. In today’s
time there are many technologies which can help in making such a system. Some technologies are “Neural
Network, Artificial Intelligence, Bayesian Network, Data mining, Artificial Immune System, K- nearest
neighbor algorithm, Decision Tree, Fuzzy Logic Based System, Support Vector Machine, Machine
learning, Genetic Programming etc”. This paper will include many surveys which will be conducted in
which people will use different techniques to make a strong system. The work will also be aiming at making
a strong detection system using libraries like numpy, sklearn and other py libraries. The problem is solved
by using a classifier which can differentiate between fraud and legit transactions based on the class and
time. The dataset contains 31 columns among which 28 columns are named as v1, v2, v3…. Due to security
purposes, 2 columns are time and amount [1]. The total amount of transactions were 283.806 with only
492 fraud cases and rest legit transactions. In today’s time there are credit cards in the market for kids who
are under 18 as well. Therefore it is important for a system to be developed for safety. Fraudsters can use
the money for many illegal practices as well. This paper will use Random Forest Classifier and Decision
tree to test the dataset [2]. The dataset is of some card holders from Europe. Keywords— Machine Leaning;
Deep Learning; Credit Card; Neural Network; E-commerce; Online shopping
PAPER : CS229
Graph Theoretic Approach to IOT Applications
Yegnanarayanan Venkataraman, Jeyaraj Jane Rubel Angelina ,Sottallu Janakiram Subhashini ,
Venkata Naveen Vadlamudi, Nagaraj P
[email protected]
ABSTRACT— Certain difficulties in the use of computing through fog and edge clubbed with the IoT
controllers in the sense of performance and memory size. So, demand increases for computing at higher
levels due to the increase in the number of IoT-oriented procedural systems required for negotiating the
problem. The transformation from the cloud poses major worries about the recognition of location proper
for nodes at the corner and the within distribution of services. It is concerned with the placement of the
selected nodes to act as hosts for the security services. This task can be visualized as a graph where vertices
stand for the devices and the edges stand for the connections between devices. Direct access from a given
vertex to a security vertex implies that chosen vertices become a dominating set of minimal cardinalities
and support heavier nodes. Here we discuss some greedy procedures that find for a given graph with allotted
weights, a dominating set obeying certain optimization criteria, and the quality of the graph partition
solution.
Keywords- Internet of Things, Graph Theory, IOT Applications, Graph Partition, Graph Vertex Set Split.
10
ISBN: 979-8-3503-4821-7
PAPER: CS241
Applications of Blockchain Technology in Modern Education System – Systematic Review
P.Chinnasamy, D. Roja Ramani, Ramesh Kumar Ayyasamy, B. Jency A Jebamani, S.Dhanasekaran,
V.Praveena
[email protected]
Abstract— The development of new technologies and also the provision of great connections are
becoming increasingly important in all aspects of our everyday situations. This advanced technology also
brings a variety of flaws, most importantly security-related problems. Distributed ledger technology,
amongst the most prominent technologies in the previous few years, is gaining traction due to its ability to
just provide privacy in a variety of domains, including logistics management, transportation, as well as
other industries. The educational industry should also make use of the important features blockchain
technology offers. This technology is now being considered by schools and universities, particularly
universities, as an approach to enhance the teaching-learning process and encourage the participation of all
stakeholders including such undergraduates, professors, and even family members. E-transcripts, digitized
certificates and credentials, data storage, plus access control will all benefit from it. This detailed research
analysis discusses the adaptability of blockchain technology in the education sector.
Keywords- Blockchain, Education Sectors, Digitized Certificates, Secure Data Storage, Access Control
PAPER: CS242
BookChain: A Secure Library Book Storing and Sharing in Academic Institutions using Blockchain
Technology
P.Chinnasamy, Anupriya Elumalai, Ramesh Kumar Ayyasamy, S.P. Kavya, S.Dhanasekaran,
Ajmeera Kiran
[email protected]
Abstract— By exploiting wireless operators, contemporary bookcrossing allows people to share books
conveniently and hastens the spread of knowledge. Nevertheless, the widespread use of smartphone
bookcrossing has been seriously hampered by the lack of tracking, resulting in missing and disappearance
of textbooks, additionally, it leads to organizational inefficiency. Hence, we suggest using the emerging
blockchain technology to exploit the advantages of its immutability and encrypted smart contract in order
to overcome these difficulties. In this research, we propose a BookChain, a blockchain-based intra-campus
book-sharing system that is verifiable and effective. BookChain securely records all of the book's
interacting entities on blockchain because every consumer could retrace the book's renting activity, hence
lowering the risk of a book being misplaced. Additionally, BookChain promotes the use of smart contracts
to streamline the distribution of books with very little human interference, maximizing productivity. The
experimental findings demonstrate the efficiency and affordability of the suggested system even with a
large number of concurrent participants.
Keywords- Blockchain, Security, Access Control, Smart Contract, Blockchain based Sharing, Secure
Library
PAPER: CS243
E- Governance Services using Artificial Intelligence Techniques
P.Chinnasamy, D. Tejaswini, S.Dhanasekaran, K. Ramprathap, Lakshmi Priya K,Ajmeera Kiran
[email protected]
Abstract— In a wide variety of fields, AI technology (AI) has lately evolved to state-of-the-art levels. It
still has a lot of obstacles to overcome before it can be ready for e-government applications, including
setting up the systems and interacting with residents online. In this research, we tackle the problems with
e-government systems & suggest a framework that makes use of AI technology to improve and streamline
e-government operations. We first design a model for the administration of egovernment information
11
ISBN: 979-8-3503-4821-7
resources in particular. Second, we frequently create a variety of deep-learning algorithms with the
intention of transforming numerous e-government services. Third, we tend to suggest an intelligent e-
government platform architecture that facilitates the emergence and execution of AI egovernment
applications. Our overriding objective aims to advance this level of e-government services by using reliable
AI approaches in order to shorten processing times, reduce costs, and raise public satisfaction.
Keywords- DDOS attacks , Dynamic path based approach, Flooding Attacks, Security, Networks
PAPER: CS244
BDDoS: BLOCKING DISTRIBUTED DENIAL OF SERVICE FLOODING ATTACKS WITH
DYNAMIC PATH DETECTORS
P.Chinnasamy, Sai Devika, Balaji V, B. Jency A Jebamani, S.Dhanasekaran, Ajmeera Kiran
[email protected]
Abstract— A distributed environment's version of a huge Denial of Services (DoS) attack against a
potential service is referred to as a Distributed DoS (DDoS) attack. Adversaries aiming for this kind of
attack make a consistent effort to exploit software bugs in computers that are utilized to attack a target
server. Zombies are computers that unwittingly help an attacker, and the attacker retains a malicious
application known as an agent on all these zombies. While occasionally solutions are developed, the
attackers are advancing their methods for launching DDoS attacks. Having defenses against both the
present and future DDoS variations is essential in this circumstance. However, managing such massive
threats calls for a thorough understanding of range and detection systems. It provides information about
DDoS attack terminology, attack kinds, and potential defenses. Assaults on and defenses against cloud
computing are also covered. If a DDoS attacks & its extent are understood, new measures for the well and
impending DDoS attacks can be created. Keywords- DDOS attacks , Dynamic path based approach,
Flooding Attacks, Security, Networks
PAPER : CS247
Adaptive Traffic Control System Using YOLO
Sirphy S, Dr. S.Thanga Revathi
[email protected]
ABSTRACT The fact is that, the population and numbers of vehicles on the road are increasing day by
day. With increasing population and hence number of vehicles, traffic jam is becoming one in every of the
crucial issues. The effect of hold-up consumes the time of the drivers and to compensate for the time wasted
at the traffic signal the drivers drive the vehicle in over speed and violate the traffic rules. The rash driving
may also have a chance of meeting accidents. Also, the effect of the hold-up of the vehicle on the road for
a long time increases air pollution. One of the factor that affects the traffic flow is the traffic signal
management system we use today. To beat the issues of traditional control systems there's a shift in adaption
to an Adaptive control system. The Adaptive control System (ATCS) could be a traffic management
technique that modifies the timing of traffic signals supported the traffic density. The proposed system uses
image processing and AI to detect the vehicles present on the road by using traffic cameras at the traffic
junction. The signal switching algorithm is based on the real-time traffic density calculated by the live
images acquired by the traffic cameras to scale back the traffic congestion.
Keywords -- Traffic light control system, Traffic management, Adaptive traffic control systems, image
processing.
12
ISBN: 979-8-3503-4821-7
PAPER : CS259
Cloud Computing Based (Serverless computing) using Serverless architecture for Dynamic Web
Hosting and cost Optimization
Rakesh Veuvolu, T.Vignesh, Anirudh Suryadevar, Nikhil Reddy Avthu
[email protected]
ABSTRACT— Serverless is a cloud-based code execution model where cloud carrie copes with servers
and computing useful resource control instead of builders. There are not any virtual machines or bodily
servers: they're deployed routinely inside the cloud by companies. Cloud carriers cope with provisioning,
keeping, and scaling the serverless architecture. What’s more, the serverless structure permits launching
apps as wished: you don’t pay for ‘constantly-on’ server components to run your app when it’s no longer
getting used. Instead, on a few occasiotriggerers app code, and the resources are dynamically allocated for
that code. You forestall paying as quickly as the code is carried out. So, in a nutshell, serverless architecture
is a manner to construct your cloud-primarily based software without coping with infrastructure. It
eliminates the want for ordinary duties like protection patches, ability control, load balancing, scaling, and
many others. Still, serverless does not mean there are no servers in any respect. The time period is truly
elusive. Servers are definitely removed from the app development seeing that they are managed by the
companies.
Keywords -- Traffic light control system, Traffic management, Adaptive traffic control systems, image
processing.
PAPER : CS260
Obstacle detection and navigation support Using SMART Shoe for Visually Impaired Person
Vijay Kumar Burugari, Kanmani P, Thariq Hussan M.I, Prabha Selvaraj
[email protected]
ABSTRACT-- Transportation and mobility is a great challenge for visually challenged people. As the
world becomes more sophisticated by leveraging the modern innovative technologies, the same can be
tapped to aid the navigation of the blind people. The most conventional way is to use a walking stick or
using a dog for guidance. Thanks to the advent and vast deployment of sensors in various applications. The
next era in the deployment of sensor is its integration in Internet of Things (IoT) where intelligence can be
inculcated in every object on the earth. There are many objects moving around like vehicles, crowds etc. It
is hard to maneuver in these without colliding for the visually challenged people. The Visually Impaired
Person (VCP) need a tool to help them navigate through the roads as there are many uncertainties for them
to be calculative. The recent growth technology can help many people who are visually challenged and
those who need assistance with their day to day activities. Our product "Smart shoe" helps the blind to
navigate in an efficient way along with the crowd. This work aims to provide efficient and cheap
alternatives that can be affordable and efficient to support blind people navigation.
KeyWords— Blind Stick, Arduino, Ultrasonic Sensor, Gyroscope, obstacle detection
PAPER : CS261
Online Medical System Using ElGamal
SHIVAM KUMAR GUPTA, Dr. S. THENMALAR, S. T. SUBHA
[email protected]
ABSTRACT— Cost-effective and secure sharing of individual health records is now possible through the
use of cloud services in the healthcare sector. Storing sensitive health information in the cloud raises
concerns about its security, necessitating an assessment of the best means of protecting this information.
As a result, we propose a strategy to facilitate the exchange of private health data in the cloud, which we
13
ISBN: 979-8-3503-4821-7
refer to as Enhanced Security Services. These systems ensure the confidentiality of PHRs while also
protecting the patient's right to exercise autonomy over their own health information. Patients encrypt their
data and store it on remote servers, allowing multiple parties access to specific sections of their medical
records. Key pairs and re-encryption keys are initially generated and stored on a semi-trusted proxy and re-
encryption server (SRS). This approach is safe because it uses both forward and backward access control
and guards against insider threats. Using High-Level Petri Nets, we formally investigate and verify the
efficacy of this strategy (HLPN). An evaluation of the personal health data methodology's efficiency
suggests it can exchange data safely via the cloud.
Keywords- Access restriction, data encryption, privacy in healthcare, patient autonomy
PAPER : CS265
Personalized Secure E-Identity Card Using Block Chain
Vijay Kumar Burugari, Kanmani P, Prabha Selvaraj, Varun Potluri
[email protected]
ABSTRACT— Blockchain technology is a new technology in digital platforms for making transactions
in a decentralized manner. Blockchain is used for last some year for bitcoin. The preceding block’s
cryptographic hash, a timestamp, and a transaction are all included in every block in the blockchain. As a
result, it is noticeable for security on an identity card. Implementation of blockchain technology for digital
identity card maintains a secure network for the user. Identity card has personal information and some
important information included. Individual person identity cards are stored in file with encryption method
which will hacked by hackers. So a system using blockchain technology for identity card is proposed. In
distribute file system, data security for information exchange and storage is provided using blockchain and
IPFS (Interplanetary File System) for storing and communicating data securely.
Keywords- Blockchain, Identity card, IPFS server, Ethereum
PAPER: CS266
AUTOMATED YOUTUBE VIDEO TRANSCRIPTION TO SUMMARIZED TEXT USING
NATURAL LANGUAGE PROCESSING
Nagaraj P, Muneeswaran. V, Rohith B, Sai Vasanth B, Veda Varshith Reddy G, Koushik Teja A
[email protected]
Abstract— This paper proposes video transcription using python language. With the adding quantum of
videotape data generated every day, it has come important to epitomize the videos for faster reclamation
and quick surfing of vids so that druggies can select the more applicable videotape for viewing as per their
demand. The significance of videotape summarization lies in the fact that it helps in effective storage and
allows quick browsing through a large number of videos. We propose a system to induce summaries for
videos by utilizing the audio element. Our videotape summarization approach involves generating the audio
paraphrase, using speech recognition, if it isn't readily available. It's grounded on assigning scores to rulings
in the reiterations and opting for the bones with the loftiest scores. Also from the original videotape, parts
corresponding to the named rulings are uprooted and intermingled to gain the final summary.
Keywords—: Summarization, Speech Recognition, Transcription, Captions, NLP
14
ISBN: 979-8-3503-4821-7
PAPER : CS267
PHISHING WEBSITES DETECTION USING MACHINE LEARNING
Saksham Singh, Suseela G
[email protected]
ABSTRACT— In order to prevent criminal activity and data theft, detecting phishing attacks has become
a pressing and difficult challenge in recent years. Phishing is a form of informational blackmail in which
the attacker (or "phisher") attempts to gain unauthorised access to private information about Internet users.
Information such as usernames, passwords, account numbers, online banking credentials, and credit card
numbers is the general information collected by the phishing attacker. Therefore, protecting the user data
and the system from malicious phishing attacks is crucial. To this end, numerous strategies for spotting
phishing scams have been devised in canonical works. Current methods suffer mostly from high time
complexity and decreased accuracy. To address these issues, the proposed work develops new security
methods that aid in the accurate identification of phishing assaults on websites. The phishing attacks are
identified using a set of various features and a classification method
Keywords— phishing, attacks, machine learning, BILSTM+RNN , Security.
PAPER: CS268
CRYPTANALYSIS OF AN ANONYMOUS AND TRACEABLE GROUP DATA SHARING IN
CLOUD COMPUTING
Dinesh Babu K M, Dr. L.N.B.Srinivas
[email protected]
ABSTRACT -- Cloud computing is a relatively new service that enables users to remotely store data. Data
privacy and security are two of the biggest concerns for people who use cloud services. In addition to being
vulnerable to attacks launched from within the cloud itself, the security measures currently in place on the
cloud platform as a whole are inadequate. Scheduling and data integrity issues are two additional challenges
for current schemes. Additional work is needed to enhance the effectiveness of existing technical auditing
alongside a variety of methods. However, this causes issues with privacy and the timeliness of data. In this
paper, we schedule data using a first-come, first-serve (FCFS) mechanism, which boosts system efficiency
and allows for parallel processing. Also, the Crisscross Advanced Encryption Standard (CAES) is
implemented for added safety. This method employs crisscross AES, which adds another layer of security
to the cloud by way of the grid. Using crisscross AES and Collocate FCFS of supremacy elements, this
proposed method improves the system's efficiency and security. Public Integrity Auditing Group Signature
and Secure Hash Algorithm (PIAGS-SHA2) is proposed for secure data integrity auditing. An additional
layer of encryption is used in this method to ensure confidentiality when storing data in the cloud. Security
is bolstered by breaking up a single file into multiple blocks and storing them under different names,
making it more difficult for would-be intruders to track down the original data.
Index Words : Privacy, integrity, scheduling, auditing, data partitioning, and anomaly detection
PAPER: CS269
Visual Cryptography for secure Banking Application
Uzma Aziz, Dr.M.Anand
ABSTRACT -- The goal of Cryptography is to send information securely while hiding it within another,
more obvious message. It's a method of concealing information by placing it within a seemingly unrelated
media type. Cryptography makes use of audio, video, image, and text in order to conceal data. Using the
lowest-bit-rate information in otherwise noisy files, multimedia Cryptography is a form of steganography
that is both highly secure and difficult to decipher. The images, which contain a wide range of colours, are
not particularly reflective. Cryptography is used to conceal cryptographic systems within image files using
a variety of methods. Phishing is an attempt by an individual or organisation to get sensitive personal data,
15
ISBN: 979-8-3503-4821-7
including passwords and credit card numbers, from unwitting victims in order to commit fraud, steal
identities, and make money. Users of online banking and e-commerce are increasingly having trouble with
phishing scams. As a result, determining whether a computer that is linked to the internet may be regarded
secure or not is essentially impossible. Therefore, to address the issue of phishing, we are employing an
anti-phishing framework based on visual cryptography. Here, visual cryptography is used to develop an
image-based authentication system
PAPER : CS271
Banking Data Migration from On-Premise to Cloud
Mahesh Gogulapati, Krishna Reddy , P.Chinnasamy, Avinash Agnihotri , Twinkle, Thondickal,
Anoop Aravind
[email protected]
ABSTRACT— Data Migration is a multi-step process that begins with analyzing old data and culminates
in data uploading and reconciliation in new applications. With the rapid growth of data, banking sector
constantly need to migrate data. Data migration can be a complex process as testing must be done to ensure
data quality. Migration also can be very costly if best practices are not followed and hidden costs are not
identified in the early stage.
Keywords; Data Migration; Banking , Cloud , Data Management, Methodologies, Azure data factory, Data
archive, Management Information System (MIS)
PAPER: CS272
TEXT RECOGNITION BASED ON ENCODER AND DECODER FRAMEWORK
Subhashini Peneti, Thulasi Chitra
[email protected]
ABSTRACT: Now-a-days we all are using digital technologies in all sections. Handwriting textbook
recognition is an active and utmost exploration areas in the field of image processing and pattern
recognition but, still we're using Handwriting clones converted into electronic clones to communicate and
store electronically. Through the textbook, we reuse the supplied image, rooting features, and feting it. The
training of the system to fete and classify objects takes place, as well as the creation of a bracket schema.
The system is trained using this system. Handwriting textbook recognition refers to detecting the computer
digital comprehensible. Handwriting textbook input for Handwriting sources similar as photos, paper
documents, and other sources. Occasionally it's complex to understand the mortal hand jotting as cursive
handwriting, Poor quality document/ image, different individualities have different handwriting styles and
other coffers. The main end of this design is to develop a Handwriting textbook recognition system which
is used to read scholars and lectures handwritten notes, croakers conventions, Research and Development
labs etc. A handwriting recognition system handles formatting, performs correct segmentation into
characters, and find correct presumptive of words. The use of neural networks for feting handwriting
textbook is more effective and robust. The end is to ameliorate the effectiveness of neural networks for
Handwriting textbook recognition.
Keywords – Presumptive, Pattern, Neural
16
ISBN: 979-8-3503-4821-7
PAPER: CS273
An Effective approach for preparation of News Summary using web scraping techniques
Mrs.N.Thulasi Chitra Dr.G. Sreenivasulu, Dr. P.Subhashini, Dr.Ganti Krishna Sharma, Mrs
N.Sandhya
[email protected]
Abstract: Due to the Internet's quick development and far reaching use, network media has turned into
another window for people to grasp the rest of the world. News is a medium through which people might
find out about their environmental factors, yet a huge number of reports are distributed consistently on the
Internet. A major need in individuals' lives is to sort out some way to get the news things we want from the
web rapidly and definitively. This strategy tries to assemble news from specific sites and give it to guests
in a reasonable and compact way. Clients might utilize specific watchwords to find news that they are keen
on, which takes into account personalization. The viable advantage of this technique is that it gives quick,
proficient, and simple admittance to news that individuals care about, need, and are keen on.
Keywords—News collection, Crawling, Scrapy, Django
PAPER : CS280
Detection Of Cyber Attack Using Artificial Intelligence
NITHISH KRISHNA AG, SAVEETHA D
[email protected]
ABSTRACT— Due to the integration of physical processes, computer resources, and communication
capabilities, cyber-physical systems (cps) have achieved notable advances in many dynamic applications.
However, these systems are seriously threatened by cyberattacks. In contrast to failures brought on by cyber
accidents in physical systems, cyber attacks take place subtly and purposefully. Particular of these attacks,
also known as spoof attacks, distort data or feed false information into the system by injecting it into sensors
or controllers. They can also do this by hacking some cyber components. If the system is not aware of these
attacks, it will not be able to recognise them, which could lead to performance issues or complete system
failure. As a result, algorithms must be modified to recognise this kind of assault on these systems. It should
be highlighted that because the data produced by these systems is produced in such enormous quantities,
with such variety, and at such a rapid tempo, it is crucial to apply machine learning techniques to simplify
data processing and evaluation and uncover hidden patterns. The CPS is modelled in this study as a network
of agents that move together, with one agent serving as the leader and the others being under his or her
orders. The approach suggested in this paper uses a deep neural network’s structure for the detection phase,
which must alert the system to the attack’s presence in its early stages. In order to isolate the offending
agent in the leader-follower system, robust control techniques have been studied. The reputation algorithm
is used by the control system in the presented control approach to isolate the misbehaving agent following
the attack detection phase using a deep neural network. Deep learning algorithms can detect threats more
effective than the traditional approaches, according to experimental studies, and they can make
cybersecurity easier, more proactive, less expensive, and significantly more successful.
Keywords- Directory-based push operations, big data analysis, parallel coordinate analysis, pattern
assessment, security threats.
PAPER : CS281
USING HONEY POT TECHNIQUE RANSOMEWARE GET DETECTED
Solomon Raja, Dr.K.Venkatesh
[email protected]
ABSTRACT— Ransomware is the challenging threat it will stop the user from their activity so the user
can’t access the data of the system. Ransomware is the challenging threat it will stop the user to do their
17
ISBN: 979-8-3503-4821-7
activity .When a ransomware is injected to a system all the utility files and the system files will be encrypted
where the user cannot work with and access with the system data. To prevent the ransom attack we need to
implement the honey pot techniques. The honey pot is the trap for the attacker it will capture everything
that enters into the system. If the packet that enters into the system is by some anonymous user and if the
user attacks the system, we can use this honey pot to trap the packet. When there is a ransom ware attack ,
the malware usually tries to encrypt the data. An honey pot is set to find if any data are encrypted in the
system. If the honey pot finds any data is been encrypted then notifies and if the honey pot fines the
encryption is started by malicious activity, then it will cut the network connectivity of the system from the
network. In this paper we have analysed different honey pot techniques. further we are going to create s-
honey pot to get prevent the attack from the ransomware.
Keywords- Honeypot, Ransomware
PAPER : CS283
COMPREHENSIVE SURVEY ON INTRUSION DETECTION SYSTEM USING DEEP
LEARNING TECHNIQUES
Jerome Shurjith D, Meenakshi.K
[email protected]
ABSTRACT: Intrusions into networks are getting increasingly common, unpredictable, and sophisticated
along with the ever- expanding size of the internet. Intrusion detection in a network of this size is a pressing
and difficult issue. Current security tools are sometimes insufficient to tackle concerns connected to
unlawful users, reliability, and providing robust network security. This is because of the increasing number
of harmful assaults on computer networks and systems. Although there have been improvements in network
security, new studies show that an increase in illicit intrusions is a serious cause for concern. As such, many
people, corporations, governments, and enterprises view security as an extremely delicate and vital issue
that must be addressed at all times and in all places. When it comes to sniffing out network packets,
monitoring incoming and outgoing network traffic, and identifying unauthorised use and misuse of
computer system networks, a NIDS is one of the most tried, trustworthy, and powerful kinds of technology.
It can help us make sense of what’s really going on in the network. More than that, a NIDS may monitor
for and report on any signs of attacks or malicious activity. A NIDS is an absolute necessity in today’s
highly connected and heavily used computer networks. More and more people are starting to pay attention
to machine learning-based network intrusion detection because of its smart capabilities. Machine learning-
based solutions, especially those employing deep learning, are superior to rule-based approaches in
detecting new and emerging forms of network assaults.
Keywords: NIDS, IDS, KDD99, NSL-KDD, Network.
PAPER : CS284
Authentication System for University Administrative Activities
VISHNU KUMAR V, ANNAPURANI PANAIYAPPAN K
[email protected]
ABSTRACT One of the biggest problems with online learning today is current learners’ organizational
plans in conformance and user identification. The trainer must verify that the internet students who have
completed the learning process and received the academic credits are the same people who signed up for
the classes, especially for online certification and accreditation. They must also ensure that these students
successfully complete all of the duties for the online training without acting dishonestly or improperly. The
movement to and deployment of online education practises have been accelerated (in some cases
dramatically) by the COVID-19 outbreak, necessitating the creation of safe methods for authenticating
and proctoring online students. Today, a number of technologies with different levels of automation are
18
ISBN: 979-8-3503-4821-7
accessible. In this article, we provide a detailed overview of a specific solution based on the authentication
of several biometric technologies and a system of automatic proctoring. (system workflow and AI
algorithms), which incorporates elements to meet the main market concerns: High levels of automation,
low cost, ease of use, little hardware or software requirements, dependability, and passiveness for the
learner This article also covers the user’s usability and privacy perception survey results, as well as the
technological performance test of the large-scale system.
Keywords: computer vision, cloud computing, data science applications in education, correspondence
courses, authentication system, protection and data analysis
PAPER : CS285
Enhancing the Security Features in Remote Keyless Entry System RAGAVENDRA PRASAD M,
SUJATHA G
[email protected]
Abstract- We are moving toward wireless network and communications in our modern technology.
Electronic locks that may be unlocked with a wireless key that transmits a predetermined code to a receiver
are known as Remote Keyless Entry (RKE) systems or Smart Key Systems.. Wireless communications
have many disadvantages, one of which is security, despite how practical they are for modern life. Attackers
can more easily exploit the wireless RKE system.AES Algorithm is used to secure the key which
transfered.The Replay attack is continuously used attack in RKE system. The main objective of project is
to secure the key from the attack by null-fying the key which is transferred already and changing the key
pattern. The RKE will have the key deleting option which is already used if any key is transmitted which
helps in securing the key. By changing the key pattern the chance of attacking the system will be reduced
from which the new key pattern will be followed to RKE system. This methods can be implement to
increase the security of the RKE system where the possibilities of the attack can be reduced.
Keywords— Remote Keyless Entry(RKE),key pattern, AES algorithm
PAPER: CS286
Cloud Storage Security Model
Harsh Malik, Mr. K. Senthilkumar
[email protected]
Abstract—Cloud storage service has shown its great power and wide popularity which provides
fundamental support for rapid development of cloud computing. However, due to management negligence
and malicious attack, there still lie enormous security incidents that lead to quantities of sensitive data
leakage at cloud storage layer. This paper proposed a Cloud Storage Security Model. To avoid data breach
at the storage layer, the model will include two different algorithms i.e., AES algorithm and Blowfish
algorithm. The experimental results indicate that proposed mechanism is not only suitable for ensuring the
data security at storage layer from leakage, but also can store huge amount of cloud data effectively and
the Blowfish algorithm provides extra security.
Keywords-Cloud computing, data dispersion, data encryption, key management, storage security
PAPER : CS290
ECC-based Authentication for Secure TDMA in
Suchi Johari, M. Bala Krishna
[email protected]
ABSTRACT— Secure TDMA in VANET is the allocation of the time slot to authentic nodes. The trusted
authority verifies the node authenticity and notifies the malicious nodes to RSU. Malicious nodes replicate
19
ISBN: 979-8-3503-4821-7
the identity of the authentic nodes in V2V communication. Multiple time slot requests by malicious nodes
in the same frame can lead to vulnerabilities in the network. This paper focuses on secure time slot
allocation for vehicles using ECC-based authentication. Registration and key exchange, sharing of the one-
time session key by RSU with the vehicle, and secure V2V communication are considered in the proposed
ECC approach. The registered vehicles are authenticated. RSU shares the one-time session key, allocates
the time slot to the vehicle, and secures the V2V communication. The malicious nodes are identified based
on a unique ID and one-time session key generated by RSU. Simulation results are compared with existing
approaches.
Keywords— Secure TDMA MAC, Secure time slot allocation, ECC, intruder attacks
PAPER: CS293
Sentiment Analysis of Twitter Reviews using SVM, ANN and CNN Models
L. Sudha Rani, C. Shoba Bindu, S. Zahoor-Ul-Huq
[email protected]
Abstract—One of the most critical factors that an E-commerce considers when it comes to enhancing its
customer experience is the availability of feedback. Through this study, we present an approach which can
analyse the tweet characters to maximize the experience of the customers. The tweet features were mined
using the word embedding and the n-gram approach. A classification model capable of identifying tweets
in either positive or negative categories was proposed by SVM and ANN architectures. A CNN-based
model was also developed to classify the tweets. The results of the study revealed that the model that was
most accurate was the one that was used for analyzing the tweets. The results of the study revealed that the
CNN-based model was more accurate than the SVM and ANN models when it came to analyzing the
tweets. The results of the study also showed that the model was able to identify interesting associations
between the various categories of tweets.
Keywords- Sentiment Analysis, Tweets, Classification, Feature Extraction
PAPER : CS302
A PECULIAR IMAGE ENCRYPTION TECHNIQUE FOR MOBILE
Krithika R, Dr.J.Godwin ponsam
[email protected]
ABSTRACT— The upgradation in the field of mobile applications is predominantly increasing. Nowadays
mobile applications are used in various platforms on one-handled devices in addition, attackers can use
similar technology to anonymize their malicious behaviors and hide their identification of behaviors.
Thus, security is important. In this project, we are focusing on the precautionary encryption and decryption
algorithms like PNSR metric and Elliptic curve Digital signature algorithm which help us to provide
secured transmission of a personal image between the mobile stations. Based on these algorithms a defense
application will be developed. There are 4 different levels of technology that will be applied in this project
which help to improve security transmission. The first level is selecting a secret image. The secret image
will support file types like jpg,png. In the second level of security, we encode the image that we get from
the first level using an encryption algorithm. Here the image quality is measured by using PSNR metric,
the third level is finding the LSB, along with 3m (Mean,Mean,Mode) of the image to hide the message
inside the cover image. Then the obtained stegnographic image is compressed using GZIP is the final
security level. An Elliptic curve, a Digital signature algorithm is used to enhance a security process.
Therefore, this method is suggested to send a secret message through applications of special importance
across the mobile application.
Keywords— Mobile application, Image Encryption, LSB, 3m, GZIP, Elliptic curve, Digital signature.
20
ISBN: 979-8-3503-4821-7
PAPER : CS303
Deep Learning-Based Sentiment Analysis on Tamil Texts In Roman script
Abhinav Ranjan, Rahul Giri, Dr. A. Pandian
[email protected]
ABSTRACT— Sentiment analysis is the process of computationally identifying and categorizing opinions
expressed in a piece of text. It can help us determine whether the writer’s attitude towards a particular topic,
product, entity, etc. is positive, negative or neutral. This paper would present a comparative study on the
different models which could be used to implement sentiment analysis on Tamil texts. This would go a
long way in being beneficial to the public, business organizations, film producers and others to make
decisions and improvements. LSTM model would be used to find polar words in a data point , which would
be a comment extracted from a film’s trailer released on youtube and hence , sentiment analysis would be
done on tamil-english texts.
Keywords- Deep Learning, NLP, Implementation of code using LSTM, Naive Bayes, ML, Ensemble
learning, Decision trees, SVM, AdaBoost.
PAPER : CS304
Yolo: Human Detection-Based Intelligent Home Automation using IOT
Nagaraj P, Sathish S, Mathan Kumar M, Arsalan Ahmad Khan, Ahmad Mohiuddin, Syed Haroon
M
[email protected]
ABSTRACT— As the Internet of Things (IoT) has revolutionized the way we live and interact with the
world around us, the need for a fully automated home is greater than ever. Home Automation allows for
the control of electronic devices in your home through the internet, eliminating the need for manual
intervention, which makes life much more convenient. In this paper, we present a model for implementing
home automation (controlling and monitoring remotely). We also incorporate sensors based on the specific
automation needs and utilize the YOLO algorithm to increase efficiency by suggesting the best usage
schedule based on user patterns. Our proposed approach for intelligent home automation integrates the
YOLO object detection algorithm, which is capable of accurately identifying and locating objects in images
and videos, with IoT devices. By using YOLO to detect human presence, we can automate tasks such as
turning off lights when no one is in the room, adjusting the thermostat based on occupancy and activating
security measures when an unauthorized person is detected. We evaluate the performance of our system
using a dataset of human images and demonstrate its high accuracy in detecting and classifying human
presence. Our results show that the proposed approach has the potential to greatly enhance the efficiency
and convenience of home automation.
Keywords— Yolo, Home Automation, Human Detection, Internet of Things, Thermostat.
PAPER : CS306
Detection Of Cyberbullying based On Online Social Networks
AISWARYA RAJEEVAN, KRISHNARAJ N
[email protected]
ABSTRACT— Online bullying is affected by cyber bullying, which has grave repercussions. It is text-
based and occurs in several forms. Most social media networks’ format. The majority of the recent studies
have focused on this topic with Most of the models developed in these studies and common machine
21
ISBN: 979-8-3503-4821-7
learning techniques are scalable at once within a single social network. Deep learning-based models claim
to be able to overcome the limitations of conventional models and improve the execution of discovery by
developing methods for the identification of instances of online harassment. Despite the fact that there are
many antiquated methods available to control the occurrence, there is still a lack of demand to successfully
control the pestering. To effectively use machine learning and language preparation to filter the pestering
in the virtual environment and to prevent the barbaric result A framework is suggested to provide two
distinct descriptions of cyber bullying. Our method makes use of CNN’s creative idea for content analysis,
however the Current tactics employ a dishonest method to handle provide the arrangement with less
accuracy. A current dataset is used for testing, and when our system is compared to other available
approaches, it is discovered to provide greater precision and grouping.
Keywords- Deep learning, machine learning, content-based cyber crime, and cyber bullying.
PAPER: CS317
Closest Celestial Body Search Using KD Trees
Dr. Priya Ponnusamy P, Shabariram C P, Umayal V R, Susmeta A
[email protected]
Abstract-A Binary Search Tree is expanded into a KD Tree to handle the multi-dimensional key searches.
A discriminator makes the KD tree different from the BST. This discriminator will take branching decisions
at every level based on the key search. It can handle multi-dimensional coordinate object searching. K-d
trees enable O(klogn) lookup times for the k nearest points to some point x. This is extremely useful,
especially in cases where an O(n) lookup time is intractable already. In this work, a kd tree has been created
based on how a celestial body is positioned in relation to the sun. It aims at finding the minimum in a
particular dimension, searching for a celestial body and its nearest neighbor.
Keywords – KD trees; Binary search tree; Nearest neighbor search; Minimum body in celestial space
PAPER : CS320
Trustworthy Web Service-Business Activities (WSBA) using an Efficient Byzantine Fault Tolerance
Algorithm
G. Muneeswari, Kanmani P, Vijay Kumar Burugari , Prabha Selvaraj
[email protected]
ABSTRACT— In order to provide the services related to coordination, all the stipulated Business
Activities (WS-BA) rely on the Byzantine Fault Tolerance (BFT) based lightweight mechanism for
trustworthiness. The main objective of this paper is to provide security and less overhead to the initiator.
The ultimate idea of the proposed work is to deploy a trustworthy third party services for the entire coupling
and coordination services for the multiple business activities of an organization or enterprises. It can also
be executed using the cloud services which will be much easy for large scale companies. In this paper we
use BFT algorithm which handles lightweight ordering of sources and provides a standard way of
communication between the client and the actual system coordinator. It also deals with Web Services-
Business Activities-Initiator service which resolves some issues related with the security to the initiator,
coordinator and participants. We also use timestamp or nonce to provide security to the coordinator and
the initiator during the request and response between them.
Keywords- Initiator, Coordinator, Business Activity, Web Services, Registration, trustworthy.
22
ISBN: 979-8-3503-4821-7
PAPER: CS321
Industry 4.0 Digital Technologies and Information Systems: Implications for Manufacturing Firms
Innovation Performance
Ramesh Kumar Ayyasamy, Farhan Bashir Shaikh, Nur Syadhila Binti Che Lah, Shadab Kalhoro,
P. Chinnasamy, Saraswathy Krisnan
[email protected]
Abstract— Cloud computing, big data analytics, and the internet of things are just a few of the recent
technological marvels that have surprised organisations, particularly when it comes to managing
innovation. The current study investigates the impact of the internet of things, big data analytics and cloud
computing as digital capabilities on information systems and firms' innovation performance via a mediating
role of dynamic and innovation capabilities. International entrepreneurial orientation has been taken as a
moderator in this study. The study uses a quantitative method research approach conducting survey
questionnaires from manufacturing SMEs in Malaysia. The conclusions of this study could aid the industry
in identifying, assessing, and strengthening its skills in the ever-evolving technology trends, allowing them
to produce new business models and innovate through information systems. People will have more
awareness about necessary technological advancements and economic and job options as a result of their
growth, and their social status will improve
Keywords- IoT; Big Data Analytics; Cloud Computing; Innovation; Industry 4.0
PAPER : CS322
Classification of Influenza from Chest Radiography using Deep Neural Architecture
T Jemima Jebaseeli, D Jasmine David, S Suriya
[email protected]
ABSTRACT— Influenza is a deadly disease that is caused by microorganisms that affects the lungs. A
survey shows that there are about 20% of the people are affected by influenza. The early diagnosis of
influenza helps the individual to speed up the healing process. Early disease identification in chest x-rays
is critical for diagnosing influenza. To detect the sickness, the proposed method employed the
Convolutional Neural Network (CNN) approach. Deep learning algorithms for disease diagnosis have
received a lot of attention. The statistical data indicate that utilising pre-trained CNN models and supervised
classification approaches to identify chest x-ray images to detect pneumonia (influenza) is clinically
helpful. Chest x-ray pictures from the kaggle dataset were tested for pneumonia. The studies are subjective
for several reasons, including the presence of infection, which might seem foggy in chest x-ray images or
be misinterpreted for other illnesses. This system may assist physicians with reliability and usability issues
that come up always when working through medical images. As a result, the proposed methodology is used
to make an early response regarding pneumonia, and also it helps the radiologists.
Keywords- pneumonia; chest x-ray; vgg16; deep learning; CNN; Convolutional Neural Network.
PAPER : CS327
Multimodal Deep Neural Network for Diabetic Retinopathy Detection
Vijay Ravichander , Aditi Mittal, Nithyakani P
[email protected]
ABSTRACT— An eye condition known as diabetic retinopathy (DR) destroys the retinal blood vessels
and impairs vision. Normal, mild, moderate, severe, and PDR are the five stages of DR (Proliferative
Diabetic Retinopathy). The risk of significant vision loss can be reduced by nearly 90% with early diagnosis
and treatment of DR. A highly skilled practitioner typically diagnoses DR by looking at the retina's vivid
fundus pictures. Diabetic Retinopathy and its various stages can be identified automatically utilizing
snapshots of the retina with various computer-vision-based techniques. These techniques however cannot
classify the early phases of DR. This study employs a hybrid deep learning architecture using Inceptionv3
23
ISBN: 979-8-3503-4821-7
and Inception – ResNetv2 to diagnose all five phases of DR while maintaining high accuracy levels for the
early stages. The proposed system consists of two steps. First off, it anticipates the early phases of DR
(normal, mild or higher). The technique will further detect for stage 3, 4, and 5, which are moderate, severe,
and PDR, if the third stage is first anticipated (Proliferative Diabetic Retinopathy). The initial and final
stages of DR may both benefit from this methodology's ability to maintain strong performance.
Keywords- Diabetic Retinopathy, Deep Learning, Convolution Neural Network, Fundus Images
PAPER: CS329
Design of secure QR payment system using Visual Cryptography method
Dr. Praveena Narayanan, Yadlapalli Lokesh, P.Charitha, G.Chethan Teja Kumar, B.Bhavya,
S.Hemalatha
[email protected]
Abstract— In this paper, we will discuss how a secure link-sharing system based on QR codes was created
and put into operation. In recent years, QR codes have grown in popularity since they expedite link sharing
and provide users with the greatest level of ease. QR-based online systems, as easy as they may appear, are
subject to a variety of assaults. As a result, link sharing must be safe enough to ensure each process, integrity
and secrecy. The link sharing system must also guarantee each transaction's sender and receiver's
legitimacy. The suggested QR-based system in this paper is secured using visual cryptography. The visual
cryptography is carried out using a web application that is part of the proposed system. Users can exchange
URLs through QR code using a simple and user-friendly interface provided by the app.
Keywords - Visual cryptography, Quick Response code and E-Payment systems.
PAPER : CS332
Convolution Neural Networks based lungs disease detection and Severity classification K.Soni
Sharmila, Dr.Thanga Revathi S , Dr. Pokkuluri Kiran Sree
[email protected]
ABSTRACT— Classification is a crucial component of Computer Aided Diagnosis (CADx) systems. This
phase comprises the extraction of features. Deep features have emerged as a new topic of study in numerous
disciplines, including medical imaging. However, these works contain flaws, such as excessive
classification, and do not reflect the real world. This paper provides an overview of deep learning for
detecting lung illness in medical photos. In the past five years, just one review article has been published
on deep learning for lung illness diagnosis. We investigate utilizing deep learning to detect and categorize
Convolution neural network (CNN) numerous lung illnesses from chest X-ray images. We developed a
pipeline for segmenting chest X-ray (CXR) images before classification and compared the performance of
our framework to that of existing techniques. To recover lung characteristics, the Binary Spotted Hyena
Optimizer (BSHO) was used in this study. We demonstrated that simple models and classifiers, such as
shallow CNN, can compete with complex systems. Furthermore, we validated our method using publicly
available lung datasets from Shenzhen and Montgomery and compared its efficacy to that of existing
methods. Despite having fewer trainable parameters, our technique outperformed the top-performing
models trained on the Montgomery dataset in terms of accuracy. In addition, although being
computationally cheaper, our CNN-BSHO model performed nearly as well as the top solution on the
Shenzhen dataset. This research employed four classifiers, including Support Vector Machine (SVM),
Nave Bayes, Random Forest, and Visual Geometry Group (VGG). Using CNN-BSHO, an accuracy of
98.324% was reached.
Keywords- Convolution Neural Network, BSHO, Lung cancer, Chest X-ray images, Machine learning
24
ISBN: 979-8-3503-4821-7
PAPER : CS336
Optimizing Virtual Machine Downtime in Live Migration using the Least Frequently Used Service-
based Pre-copy Algorithm
Mr. C. P. Shabariram, Ms. R. Prema, Dr. N. Shanthi, Ms. R. S. Sowmiya Priya, Dr. P. Priya
Ponnusamy
[email protected]
ABSTRACT— Cloud computing has been considered the most reliable and safe technology among many
cloud users. Virtual machine migration is a crucial component of cloud computing technology. A Virtual
machine is transferred from one host machine to another during the migration process, which is
accomplished by repeatedly copying the memory pages available in the virtual disk. The time required for
the movement of secondary storage, main memory, process states, and network connectivity, as well as the
maintenance of their accuracy, must be considered for the efficient migration of virtual machines from one
physical host to another in the cloud service provider. The most robust mechanism for system maintenance,
resilience system, balancing the load of the service provider, and power-saving is the live migration of
virtual machines between the nodes in the data center. So many services are running in the virtual machines
during live migration; thus, it needs to take place with the least amount of downtime possible to prevent
service degradation. The memory pages in the virtual storage must be copied to the other host using the
least frequently used service-based pre-copy algorithm to optimize the downtime. The effectiveness of the
suggested solution was evaluated in comparison to OpenNebula and OpenStack, two cloud computing
platforms. The least frequently used service-based pre-copy algorithm reduces the amount of virtual
machine downtime during live migration.
Keywords- Cloud Computing, Virtual Machine, Live Migration, Downtime, Least Frequently Used Service,
Pre-copy, Total Migration time, OpenNebula, and OpenStack.
PAPER : CS339
Age and Gender Prediction using Adaptive Gamma Correction and Convolutional Neural Network
Anweasha Saha, Nithish Kumar S, Nithyakani P
[email protected]
ABSTRACT— The classification of age and gender has drawn increased attention recently because of its
significance in creating user-friendly intelligent systems. In the domains of image processing and computer
vision, determining age from a single facial image has proven a challenging job. Convolutional Neural
Network (CNN) based techniques have been frequently adopted for the classification problem in the recent
past because of their precise results in facial analysis. This study incorporates an end-to-end CNN approach
with the addition of a key pre- processing step for image contrast enhancement, which was done via an
adaptive gamma correction technique to produce precise gender and age group classification of real-world
faces. The complete feature extraction and classification processes are included in the two-level CNN
architecture. The feature extraction task pulls features that are related to gender and age while the
classification assigns the facial photographs to the proper gender and age group. The proposed network has
been trained and tested on the Adience (original) dataset. The results of the experiment seem to back up
the claim that the proposed model is better at classifying people by gender and age when the Audience
benchmark for classification is used.
Keywords- Age Identification, Facial Images, Gender Classification, Convolution Neural Network, Deep
Learning
25
ISBN: 979-8-3503-4821-7
PAPER : CS345
Gait speed based individual recognition model using deep 2-D convolutional neural network
Abhay Lal, Nithyakani P
[email protected]
ABSTRACT— GAIT represents the pattern in which a person walks. We have four types of walking
conditions: normal walking, slow walking, fast walking, and normal walking with a bag. GAIT analysis
has been beneficial to doctors to analyse the patterns of walking to detect pains and diseases. In this work,
we will compare the classification of the four types of conditions of GAIT using 2-D Convolutional neural
networks. The CASIA C dataset has been used to train the 2-D Convolutional neural networks that have
images captured using thermal infrared night imaging. Using transfer learning, we tested the performance
of various CNN algorithms like DenseNet201 and MobileNetV2. The various algorithms provide us with
information that the GAIT C dataset is best for the custom CNN model created and we have compared all
the results in this paper.
Keywords— Gait walking pattern; Gait analysis; deep convolutional neural network; deep learning.
component;
PAPER: CS346
E-Coating Ultrafiltration System Maintenance using Machine Learning Techniques
D Anusruti, Dr.PCD.Kalaivaani,S Dhanvarsha, R.Ashwin, S Gowtham, S Gokulsundar
[email protected]
Abstract— Ultrafiltration process is one of the important processes in e-coating of metal parts. It is
important to maintain and improve the performance of ultrafiltration membrane (UF membrane). This UF
membrane should not be degraded under any situation, if this happens then it might lead to the flooding of
coating fluid all over the metal parts. Hence it has to monitor properly. This will also lead to the wastage
of coating fluid, wastage of materials and maintenance cost will also be high. So to avoid this, the workers
have to monitor it on periodically basis. During e-coating in electrophoresis painting plant there may occur
fault in filter and excess of fluid will flood in the material to be coated. The filter used in the ultrafiltration
subsystem has to be monitored manually each time by the worker. Sometimes, if the worker did not monitor
properly, it might lead coating to inappropriate places, it will cause wastage of material and fluid and also
it takes time for clean-up and to change after fault occurs. Hence, the timely detection of faults is very
important to prevent damage.In order to prevent these kinds of situations, a machine learning model is
developed which predicts the flow meter readings beforehand. It uses an ensemble learning algorithm
known as XGBoost which is one of the more powerful tools for prediction.
Keywords- Ultrafiltration, UF membrane, XGBoost, Coating, membrane, Introduction
PAPER: CS349
Unsupervised Online Video Object Segmentation
R. Mahima, Dr. M. Maheswari, E. Priyanka, C. Praiselin, K. Sanjitha
[email protected]
Abstract— Video segmentation refers to reading video photos and segmenting them into areas of interest.
The unsupervised video segmentation performs critical position in huge style of packages from item
identity to compression. The unsupervised online video object segmentation structure is proposed with the
aid of using implementing the movement property, transferring in a concordance with a standard item for
segmented areas. By incorporating notable movement item proposals and detection, a pixel smart fusion
policy is advanced efficiently to locate and do away with noise which include dynamic heritage and desk
bound objects. Furthermore, with the aid of using leveraging the received segmentation from without delay
26
ISBN: 979-8-3503-4821-7
previous frames, an ahead propagation set of rules with hired to address unreliable movement detection
and object proposals.
Keywords- Unsupervised Online Video Object Segmentation (UOVOS), Digital Image Process (DIP)
PAPER: CS351
A Systematic Review on classification of Cyber Attacks and its Prevention techniques to improve
Cyber Security
Dr.A.P.Nirmala, Dr.Asha, V. Dr.B.Nithya Ramesh, Chandana.K, Chandana.G.R, Arshe Alam
[email protected]
Abstract - Cyber security is an important contemplation for both information technology and Internet
services. People must identify the significance of various kinds of risks that occur in the online world.
Improving cyber security and safeguarding serious information are important phases to the nation's security
and economic well-being. Cyber security issues have evolved into national security concerns. When users
consider cyber security, they immediately think of “cybercrime”, which is on the growth. Various
governments and businesses are taking numerous steps to combat cybercrime. This paper mainly
emphasises cyber attacks and different kinds of cyber-attacks and their prevention techniques to improve
cyber security. Classification of cyber security to achieve the goals such as confidentiality, integrity, and
availability. It also provides a detailed view of sectors affected by cyberattacks and trends influencing cyber
security. Cyber incidents highlight the importance of staying up-to-date on global cybercrime trends,
especially personal computing devices and mobile devices.
Keywords – Cyberattack, Cyber Security, Intrusion Prevention Systems, Malware Scanners, Hacking,
Auditing and Logging.
PAPER: CS354
Anti-Accident Mechanism to Detect Driver Drowsiness Integrated with Alerting Mechanism: An
IoT-based Model
Srujan Katukam, Sai Vinay Potluri, Sunell Dutt Abburi, Rahul Cherukuri, Sreeka Parinishta C,
Mohith Peddineni, R. P. Ram Kumar, Sanjeeva Polepaka
Abstract—According to the report released by the National Safety Council (NSC), about 1,00,000
accidents, 71,000 injuries, and 1,550 deaths are caused due to Drowsiness while driving each year.
According to a study by the AAA Foundation for Road Safety, Drowsiness is responsible for up to 9.5%
of accidents and 10.8% of those involving airbag deployment, injury, or significant property damage. The
paper aims to build a real-time system to prevent accidents caused due to Drowsiness. It implements the
features of Computer Vision, Deep Learning, and Internet of Things (IoT) tools to detect Drowsiness and
alert the Driver. The various alerting techniques adopted to alert the Driver includes Alarming, Vibration,
and Water Sprinkler are implemented using IoT.
Keywords-Camera, Eye Aspect Ration (EAR), Mouth Aspect Ration (MAR), DCNN, Buzzer, Vibration
Motor, Water Sprinkler.
PAPER : CS360
Classification of Tweets Using Natural Language Processing from Twitter API Data
Nagaraj P, Muneeswaran V, Ritika M, V Saranya, N Gayathri, P V Gowthami
[email protected]
ABSTRACT— With coarse, diverse, and likely unfiltered general knowledge and the unrealistic state of
manually tagging a significant number of tweets using coaching classifiers, job sentiment analysis
classification Obtaining information for coaching children becomes a major challenge. In this article, we
present a solution to axiomatically retrieve labeled, filtered, large-scale coaching job information from
27
ISBN: 979-8-3503-4821-7
Twitter that can be used as input to a support vector machine classifier. There are various implements and
methods for performing sentiment detection and analysis on the side of supervised machine learning
algorithms. This algorithm performs categorization on a prey collection instructed on coaching job
information. Here, we utilized support vector machines (SVMs) for sentiment analysis on Weka. SVM is
a well-known supervised machine-learning technique for detecting ground polarity.
Keywords— Machine Learning, Natural Language Processing, Support Vector Machine, Naïve Bayes,
LSTM.
PAPER : CS363
AUTOMATION OF BIG MART DATA SALES USING SUPERVISED MACHINE LEARNING
TECHNIQUES
Nagaraj P, Muneeswaran V, P Naga Ramya Sri, P Amulya, Y Sushmitha, K L Saranya
[email protected]
ABSTRACT — A group of methods known as machine learning enables software system applications to
improve the accuracy of their outcome predictions. The strategy is based on data from Big Mart sales outlets
for 2013. In a variety of businesses, including retailing, shipping, manufacturing, selling, and wholesaling,
sales forecasting is crucial. It enables businesses to operate at a high level of efficiency, predict successful
sales revenue, and establish a better plan for the corporation's future expansion. In order to forecast the
right outcomes, information exploration and feature engineering are crucial. Building a prognostic model
involves taking into account many features of a dataset gathered for a big outlet as well as the methodology
used. Since these models are applied in many situations and are instructed to satisfy management
expectations, the organization's aim is reached through a variety of techniques. In this study, we look at the
results of six machine learning algorithms for forecasting sales at Big Mart and discover that one model
performs very well in comparison to other algorithms. These results will help increase sales and forecast
demand in the future.
Keywords: Machine learning, Continuous values prediction, Stacking, Sales Prediction, Linear
Regression, Ridge, Lasso Regression, Decision Tree, Random Forest, TAM.
PAPER ID: CS364

Aspect Based Sentiment Aware Word Embedding for Cross Domain Sentiment Analysis
G. Thamarai Kannan, N.A. Ponnazhagan, S. Sudhan, Gunasekar M, K.R. Prasanna Kumar
K Logeswaran
[email protected]
ABSTRACT: - Sentiment analysis is a branch of study that focuses on determining how well subjective
literature is written by examining people's attitudes, views, and sentiments toward various objects. Data
analysis and opinion word extraction from the data are difficult tasks, especially when they involve
evaluations from entirely unrelated fields. In Aspect Based Sentiment Analysis(ABSA), Aspect word
extraction and sentiment of the aspect word extraction are significant job. On these two tasks, numerous
methodologies have recently advanced. Few studies, meanwhile, attempt to separate opinion objectives and
words as pairs. In this research, we present ABWE (Aspect based word embedding) a sequence labelling
subtask for ABSA that tries to filter the appropriate sentiment words for a aspect word. To do this, aspect-
fused and context-fused sequence labelling neural network model is created. By using AFCR, the aspect
word information is effectively joined into the context words using LSTM network. The global context
representation is prepared using Bi-GRU. The encoder creates a final word representation, where the global
context information, aspect and opinion information are merged. Using the ABSA benchmark dataset from
laptop and restaurant reviews, we construct four datasets for ABWE. The results show that ABWE model
28
ISBN: 979-8-3503-4821-7
provides better results than the other baseline models. We think that our approach may be useful for pair-
wise opinion summarization as well as downstream sentiment analysis tasks.
PAPER: CS367
The Role of ERNIE Model in Analysing Hotel Reviews using Chinese Sentiment Analysis
Lee Hao Jie, Ramesh Kumar Ayyasamy, Anbuselvan Sangodiah, Norazira Binti A Jalil, Kesavan
Krishnan, P. Chinnasamy
[email protected]
Abstract— This research aims to find the best deep learning model to do Chinese sentimental analysis.
BERT's model may work well in the English language but not work in the Chinese language. English is
easier to encode in embedding space, and words are separated by space characters, unlike Chinese words,
which do not exist separate by the area. BERT Chinese model work on character level encoding is not
enough to make the BERT model adapt to Chinese NLP tasks. The BERT Model and ERNIE models are
the current state of art pre-training language models. Word level embeddings and input is needed to improve
the BERT model to perform in Chinese Sentiment Analysis. With the motivation to improve the Chinese
Sentiment Analysis, this study will explore the role of different models to propose a better version of the
ERNIE model. This study will limit the scope to improve Chinese sentiment analysis among different NLP
tasks.
Keywords-component; Opinion mining; Chinese Sentimental Analysis; ERNIE; Pre-trained models
PAPER ID: CS369
A Review of Blockchain Technology in Smart Applications: Research Issues and Open Challenges
B.Samuel , Dr. K. Kasturi
ABSTRACT-- The increasing urbanization of the worldwide people has a significant impact on societal
attitudes, particularly life expectancy, as a consequence of a variety of macroeconomic, sociological, and
environmental changes. The idea of a "smart city" creates opportunities for addressing important urban
problems. Smart cities strive to provide exceptional services to the populace, more efficiently use public
funds, and advance general welfare. Technology tools are essential to the development of ubiquitous
computing. Trustlessness, traceability, anonymity, democracy, automation, decentralization, and
confidentiality are just a few advantages of blockchain technology.The development of IoT technology and
the upgrading of intelligent devices are both facilitated by these cryptocurrency capabilities. In this post,
we will provide a thorough assessment of published literature on decentralized blockchain systems in wide
variety of applications. We start with talking about the core ideas behind distributed ledger technology
before moving on to the various fields it has been used in, like the Internet of Things, agribusiness, financial,
as well as other businesses. Finally, we examined several security issues facing the healthcare sector and
thoroughly examined blockchain technology's trustworthiness.
Keywords— Blockchain Technology, IoT, Smart Healthcare, Agriculture, Banking, Smart Cities.
PAPER ID: CS370

Empirical investigation on the security and threat mitigation of campus switches
Mooraka Sandhya
[email protected]
ABSTRACT— The OSI seven-layer model has emerged as a global consensus for safely and confident
organizational network communication while preventing other OSI layers from being familiar with the
existing layer of communication. The application, transport, and network layers have all seen extensive
research and development into solving network threats. The maintenance of Layer 2 threat mitigation is
still insufficient. Unfortunately, this means that communications can be affected if one layer is broken
without the other layers being aware of the issue. The entire network may be vulnerable if a first attack
29
ISBN: 979-8-3503-4821-7
originates at Layer 2. The port security function must be set up correctly in order to work well in an
organizational setting; otherwise, it may rapidly become more of a burden than a benefit. All assaults and
defense strategies rely on an IP-enabled switched Ethernet network. Most of these attacks become
considerably easier if shared Ethernet access (WLAN, Hub, etc.) is implemented. This paper summarizes
various potential Layer 2 attacks while keeping an eye on security issues in Layer 2. VLAN hopping, MAC
flooding, DHCP attacks are just a few examples of the network security issues caused by inadequate layer
2 hardening that are covered in this paper. Additionally, it addresses how this increases a LAN's or network
system's susceptibility to attacks in general. How to set up switch efficiently to mitigate the attacks will be
demonstrated.
Keywords- - Media Access Control Address, Dynamic Host Configuration Protocol, Virtual Local Area
Network, Address Resolution Protocol spoofing, Open Systems Interconnection, Security, attack, Layer 2;
PAPER ID: CS376

CLASSICAL REVIEW ON STOCK PROGNOSIS USING LONG SHORT TERM MEMORY
Kuncham Sri Venkata Ramya Praneetha, Perela Narayana Nithesh, Pinni Venkata Sai Sumanth
Kumar, T.Vignesh
[email protected]
ABSTRACT — The main plot of this project is to compare different methods that primarily finds the stock
prediction. There are so many machine learning and neurocomputing fields to predict the stock values.
Stock price forecasting uses machine learning effectively. Various learning methods are available,
including Moving Average (MA), K-nearest Neighbors (KNN), LSTM (Long Short Term Memory), and
ARIMA.. LSTM is a type of ANN and RNN neural networks. Because in deep learning they are capable
of storing the data in memory. But out of this Long Short-Term Memory(LSTM) is unique compares with
other. Because it is used to create long term memory.LSTM performs better in big datasets. It has more
additional space to store longer information and it stores longer period of time. While compare to other
techniques they can't store more data like LSTM does. In LSTM we use visualization method so, it is easy
to compare the data and find the accuracy value .we used apple and google company datasets to perform
LSTM .And here in the papers which we used as reference they
performed on datasets namely- Chinese stock market data, Yahoo finance dataset(900000), BSE stocks
Tick data, LM SCG, 4 year data of NASDAQ,100 stock market data of NASDAQ stocks .LSTM method
gave us best accurate value so we choose LSTM methodfrom others. The main objective of this paper is to
find the accurate value of stock market using machine learning through best method out of all available .so
we choose LSTM method.
Keywords – Long short term memory, KNN, ARIMA, Datasets, Moving Average method, RNN
PAPER: CS382
Automated E-Commerce Price Comparison Website using PHP, XAMPP, MongoDB, Django, and
Web Scrapping
Nagaraj P, Muneeswaran V, A V S R Pavan Naidu, N Shanmukh, P Vinod Kumar, G Sri
Satyanarayana
[email protected]
Abstract— Background: Every online user wants a detailed review about the product they buy or the
product they are interested in. Buyers research various websites on the internet to get the best offers on
their desired products. Purpose: We need a model which can give us the details of the products from
different websites. Method: In this paper, we are trying to utilize PHP, XAMPP, and MongoDB to build a
website to get the prices of any desired product from Amazon and Flipkart. Result: Experimental
30
ISBN: 979-8-3503-4821-7
evaluations showed that this proposed website shows the prices of the desired product from Amazon and
Flipkart with an accuracy of 96%.
Keywords—Price Comparison, Products, Amazon, Flipkart, Web Scraping, Data Extraction, XAMPP,
PHP, MariaDB, MongoDB, Django Web Framework.
PAPER ID: CS383

Enhanced Data Security using Trust Attribute-Based Encryption in Cloud Computing
T.A.Mohanaprakash, B. Prabha, A.Vinothini, G. Nalinipriya, S.Lincy Jemina , P M Suresh
[email protected]
ABSTRACT — Cloud computing helps adaptable and advantageous approach to sharing the information.
Cloud computing Plays a significant job in Society and Individuals. Information security and Data Privacy
are critical issues in large data storage environments. In Existing System Cloud figuring access is unified.
In Centralized methodology Key disseminates to every one of the clients. The Main disadvantage is
Updating of properties isn't productive and High Complexity. Secure information salvage assumes
fundamental part in all correspondence conditions so in this paper proposed Trust Attribute Based
methodology. The Proposed framework further develops the information security and Privacy. The
Attribute based methodology have different Technologies for Providing the Security to sharing the
information those are Key Policy, Cipher text Policy Attribute-Based Encryption, File Hierarchy Cipher
text strategy Attribute-Based Encryption. Trust Attribute Based methodology fundamental Concern is that
unscrambling can happen just if the Trust based worth and area fulfilled the Trust conditions and client
holds the key with Matching Attributes. Key Identification key procedure to oversee huge number of cloud
administrations in secure way. At long last our proposed approach gives answer for the information security
and information uprightness to Large Cloud stockpiling.
Keywords – Trust Attribute based encryption, Identification Key policy, Cipher-text policy, File Hierarchy.
PAPER ID: CS384

EGMPIP: Enhanced Geographic Multipath Routing using Gossip-Based Direct Neighbor Discovery
Algorithm in Wireless Sensor Networks
N.Senthilkumar, Dr.N.Revathy, Dr. T. Guhan
ABSTRACT— - Many paths between the source and sink nodes may be detected using multi-path routing
in WSNs. The reliability, energy efficiency, and other benefits of data transmission may be increased by
using many channels. The topic of multi-path routing has been the subject of several academic studies.
These multi-path routing protocols, however, are ineffective because of the time and energy needed to build
several channels. The elimination of bottlenecks and delays in data transmission is made possible by
multipath routing in WSNs. Network QoS (Quality of Service), load balancing, and communication
dependability may all benefit from multipath routing systems. To identify multipath routing in WSN, we
suggested the Enhanced Geographic multipath (EGMPIP) architecture. The proposed framework first
constructs the base station routing algorithm by computing the Neighbor node discovery (NND) function
in the network using the G-ND. Each node in the network may make use of multipath routing thanks to the
network-wide routing mechanism. In the end, simulation data are shown to validate the suggested
technique, and the findings demonstrate that the routing strategy described in this work is superior to past
methods.
Keywords – Neighbor Node Discovery, Wireless Sensor Network, Quality of Service, Multipath Routing,
Enhanced Geographic multipath.
31
ISBN: 979-8-3503-4821-7
PAPER ID: CS390

Emergency Tracking system using Intelligent agent
V.K.G Kalaiselvi, Susila N, Hariharan Shanmugasundaram, Srinidhi M, Surya Kumar R
Advaith Krishna
[email protected]
ABSTRACT— Emergency tracking is an important research aspect in todays modern era. In this current
situation, the important goal is to provide a security system for both men and women who are sexually
harassed. Women are not safe while traveling alone on lonely roads and deserted places. The study
presented here in the paper reports that 24.8% of people have experienced physical and/or sexual violence
at some point in their lives. To overcome this situation several methodology exists with integration of
multiple devices like IoT components which includes Arduino Nano, Pulse sensor, GSM Module etc. This
smartwatch connects to the smart phones thereby encouraging to secure themselves in an emergency
situation by enumerating about it to the administrators or the nearest police station by means of initiating
an ALERT Message. A device/app is designed in such a way that a smart watch and an app is connected
through BLE (Bluetooth low energy) and it retrieves the OTP as well as last received calls, thereby merges
it into a single data and then sent to official’s cloud thereby providing enhanced security. The paper presents
a proposed architecture and preliminary model and also analyses the framework with several intelligent
agents that deemed to fit for as compared to existing models.
Keywords- Intelligent system; Artificial intelligence; Smart watch; GPS/GSM; Smart phone application;
machine learning
PAPER ID: CS391

Intelligent system; Artificial intelligence; Smart watch; GPS/GSM; Smart phone application;
machine learning
K. Rajalakshmi, M. Sambath, Linda Joseph
[email protected]
ABSTRACT— The flexibility and reliability with which cloud computing can deliver a wide variety of
computation, memory, communication, and resource management services through the internet have
contributed to its rise in popularity in recent years. Increasing interest in cloud computing is due to the
growing need for reliable, affordable infrastructure. This essay largely focuses on an overview of problems
and unresolved difficulties with cloud databases as storage for cloud computing. This survey's objective is
to give a researcher a broad perspective on the problems, unresolved problems, and emerging trends in
cloud storage, with a particular focus on cloud databases. This article compares academic approaches and
methodologies to address the difficulties related to data availability, including replication, managing, and
securing data in cloud enabled technologies.
Keywords: Availability, Consistency, Cloud Computing, Cloud Database, Cloud Storage
PAPER ID: CS392

Weather Prediction using linear regression model
U Vivek Krishna, S Sudhakaran, S Sanju, R Kaladevi, Hariharan Shanmugasundaram
[email protected]
ABSTRACT— Weather prediction has been a delicate task in the meteorological department since times
due to varied reason similar as the drastically unpredicted behaviour of climate. These predictions are
important because they impact diurnal life and also affect the frugality of a state or indeed a nation. Our
model uses machine literacy algorithm to predict the weather. Our model is intended to provide accurate
32
ISBN: 979-8-3503-4821-7
weather prediction. Long- term climatic change on Earth and the goods it'll have on current and unborn
generations' lives are both unknown.
Keywords— Weather prediction, Machine literacy, Forecasting
PAPER ID: CS393

Sewage Monitoring System Using IoT
Gerlin V, Gnana Priya Y, Swetha k R , Kaladevi R, Hariharan Shanmugasundaram, Bhanuprasad
A
ABSTRACT— Since most smart cities are built with subterranean drainage systems, keeping them clean
and hygienic is exceedingly challenging. So, there is a huge possibility that this sewage may get clogged
and cause pollution to the environment. So, it very much essential to have sewage monitoring system that
could easily detect the blockage in the man hole. Sewage pipelines are very narrow and deeper pipelines
which are the major source of poisonous and hazardous gases. The accumulation of waste in the sewage
for extended periods as well as chemical processes taking place on them results in the development of
several dangerous gases with exceptionally high concentrations, including hydrogen sulfide, ammonia,
carbon dioxide, and methane. Therefore, breathing in these gases for an extended period can result in
chronic ailments and even death. In this system, to monitor sewage, IoT technology is used (which uses
gas, ultrasonic, temperature, and heartbeat sensors), to detect the presence of blocks in the sewage and also
used to measure the concentration of toxic gases present in the sewage.
Keywords— Sewage Monitoring System; IoT; Sensor; BlockDetectors
PAPER: CS395
Classification of status of fruits using Deep Learning Algorithms
Vinothini P, Varadhaganapathy S, Sanjai S, Yogeshwaran R
[email protected]
Abstract - The nutrients, vitamins, and minerals found in fruits are abundant. The fiber is also present.
Fruit classification has a wide range of uses in both industrial and agricultural environments as well as in
the service sector. For instance, it can be used in the juice industry to distinguish between fresh and rotten
fruit and may even assist some individuals in determining whether a particular fruit type satisfies their
dietary needs. A supermarket cashier may also utilize a fruit classification system to quickly and easily
determine the types and costs of the various fruits. Manual sorting, which was previously utilized for fruit
classification, is timeconsuming and necessitates the presence of a human at all times. In this study, a
machine learning-based method for categorizing and identifying five different types of fruit is described.
The method uses a dataset of 5000 photos, of which 4,000 are used for training and 1,000 for testing. The
framework is based on two distinct deep learning architectures, to be more precise. We employed a deep
learning method that has been widely applied to image recognition, such as GoogLeNet and ResNet50. In
ResNet50, where it had been trained for 10 epochs, it achieved 99% accuracy, and in GoogLeNet, 73%
accuracy.
Keywords – Fruit Classification, GoogLeNet, ResNet50, Deep Learning Algorithm, Convolutional neural
network.
33
ISBN: 979-8-3503-4821-7
PAPER ID: CS398

Imbalanced Breast Cancer Prediction Using CNN Through Multi-Net Framework
Roopa Devi E M, Sanchana Shree K I , Shanthakumari R, Shakkirun R, Rajadevi R
Poorvaja K
[email protected]
ABSTRACT— Cancer remains to be one of the most prevailing diseases in the world in which breast
cancer remains unaddressed. There are several methods for identifying breast cancer, but still, breast cancer
detection using automated algorithms remains a problem within the literature. Though several automated
algorithms are used to identify breast cancer, the one with the highest accuracy needs to be identified. CNN
abbreviated as Convolutional neural network, a fragment of a deep learning model, was created to
accurately identify breast cancer. There are two objectives in this research where the introductory step is
to look into assorted DL methods for cataloguing breast cancer images. Secondly, to identify the best
feature extraction model. The proposed system, analysis of different performance metrics results that were
obtained for the deep neural networks showed that other factors, such as transfer learning approaches and
pre-processing techniques, may have an impact on the model’s ability to achieve improved prediction of
breast cancer. Approaches like EfficientNet, Inception V3, VGG-16, and Modified VGG-19 have been
used to identify the most effective results for the ImageNet database. In this experiment the experimental
design for each model tested on breast cancer histopathology images is carefully analyzed and discussed.
Keywords— Imbalanced datasets, Convolution neural network, Multi-Net framework, Transfer Learning,
Deep learning.
PAPER ID: CS403

Classification of Breast Cancer Diagnosis Using Machine Learning Algorithms
Raghav Dwivedi, Khush Ramdev, Er. Meenu Jain, Dr. Neha grawal
[email protected]
ABSTRACT— In the developing world, cancer is one of the major causes of death. Breast cancer has the
largest share among all types of cancer among women in India [1] and worldwide [2]. Early diagnosis of
this can help the patient to increase their chances of survival, as it can provide much needed treatment on
time. This paper focuses on classification of breast cancer into benign and malignant using various machine
learning methods namely Logistic Regression, Decision Trees, Random Forest, K Nearest Neighbor,
Extreme Gradient Boosting and Support Vector Classifier on the basis of Breast Cancer Wisconsin
(Diagnostic) dataset. We propose a comparison between these implementations and evaluate them on
various parameters. This survey also analyses the effects of dimensionality reduction using Principal
Component Analysis and Recursive Feature Elimination on a few of these algorithms.
Keywords: component; machine learning, breast cancer, classification, random forest, XGB, logistic
regression
PAPER ID: CS407

LIFE-SAVING OPULENT AMBULANCE TRACKING SYSTEM
Gunavathie M A, M. Manikandan, Ashwathi A, R. Gowthamani
[email protected]
ABSTRACT— The ambulance driving system is very useful in case of emergency. This life-saving
opulent ambulance tracking system works in two ways. First, users can be contacted quickly via mobile
phones in case of emergency, and second, ambulance drivers can easily navigate to the user's location. The
phone itself already has this system app installed and is likely to do several of the necessary actions. So,
34
ISBN: 979-8-3503-4821-7
you can use this system to contact the closest ambulance driver or the closest emergency health care. We
developed this system to track your location and display both origin and destination.
Keywords- ambulance, doctor, emergency, application
PAPER: CS408
AI based speech recognition of literacy to improve tribal English knowledge
V.Saranya, Devi.T, N.Deepa
[email protected]
Abstract— Indian tribal peoples have some issues in their education. They are differentiated by their looks,
appearance and language. This kind of differentiation makes their life in trouble compared with the normal
life of other peoples. One of the main reasons for this difference is their language. Language of tribal people
cannot be understood by other peoples of India because their language is so different from other languages
and it cannot be understood and learned easily from them. Because of this they cannot go to school and do
not continue their studies as well. Teaching language is differed from their mother language so the tribal
children’s unable to understand what they are teaching and nothing learned from schools. Due to this issue,
they cannot continue their studies properly. To overcome this issue, they continue studies in their mother
language, nowadays digital technologies play an important role in Education field. Artificial Intelligence
based technology is to help tribal people education, and teaching them in their mother language. Substantial
CNN approach proposed in this paper. Feature extraction done by spectrogram and audio signal and it
performed well by this approach. This approach overcomes the limitations and performs more efficiently
than the existing approach. It produced an 85.67% accuracy range.
Keywords— Artificial Intelligence, Deep Learning, Machine Learning, CNN, Audio Classification, Speech
Recognition
PAPER ID: CS411
A CNN Based Machine Learning Scheme to Detect Lung Cancer Detection from CT Scan Images
Dr T M Saravanan, Dr.M.Jagadeesan, Dr.P.A.Selvaraj, M.Mohamed Mubarak, S.Praveen Kumar,
S.Sanjay
ABSTRACT— The major cause of mortality in today’s globe is lung cancer. As a result, respiratory organ
nodules, which are round or oval-shaped growths of scar tissues, must be identified properly and quickly
in order to prevent lung cancer from developing. As a result of critical tasks, oxygen to the human body is
provided by lungs while carbon dioxide is excreted; require early identification of malignant cells. In order
to reduce the chance of mortality, it is critical to begin therapy by detecting cancer using a CT scan. The
suggested technique involves changing the parameters, combinations that are dynamically pooled, dropout
procedures that are added, and linear unit functions that are corrected (ReLu) to build a model for detecting
lung cancer and categorizing tumors discovered in the lungs as malignant or benign. To forecast greater
accuracy, this projected method is compared to the pre-trained model VGG16. This scheme will make it
easier for doctors to give early treatment and reduce mortality.
Keywords— component; Lung cancer, CNN, VGG16 and CT scan image.
PAPER ID: CS412

Comparative Analysis of Canny Edge Detection-Based Smart Traffic Control System Using Multi-
Type Image Processing
Nagaraj. P, Muneeswaran V, K. Saiteja, D. Abhishaek Yadav, Manikanta K, S. Krishna Aditya
[email protected]
ABSTRACT— In this paper, Using the Canny Edge Detection Algorithm, we are introducing a method to
automate or regulate green traffic signal allocation time based on traffic conditions at the roadside. We are
sending the most recent traffic picture to the program to use this strategy, and the application will extract
35
ISBN: 979-8-3503-4821-7
edges. If there is more traffic, then there will be more white-colored edges, and if there is less traffic, then
there will be fewer white-colored edges in the uploaded image. Empty edges will be black with a value of
0. We can determine the total amount of traffic by counting the non-zero white pixels; then, based on that
information, we can decide how much time to give for the green signal. Green signal allocation time will
be longer if there is more traffic, and shorter if there is less.
Keywords- Edges, Canny Edge, Technique, congestion, Detection
PAPER ID: CS413

A Comparative Analysis of Retinal Disease Image Classification for OCT Using Deep Learning
Techniques
Nagaraj P, Muneeswaran V , Sunethra B, Sreeya C, Dhannushree U , Nithisaa S [email protected]
ABSTRACT— Classifying retinal disorders with optical coherence tomography (OCT) scans is crucial.
For patients to survive, these diseases must be accurately detected and classified. The analysis of retinal
illnesses is currently done by doctors regularly by looking at optical coherence tomography (OCT) pictures.
The manual diagnosing process is laborious, though. In this work, an automated detection and classification
system for retinal diseases has been developed to assist doctors in their diagnosis. The noninvasive OCT
test, which generates cross-section images of the retinal layer of the eye, allows ophthalmologists to
diagnose conditions based on the layers of the retina. As a result, it is a crucial modality for the
identification and measurement of retinal disorders and diseases. Ophthalmologists must spend a lot of time
analyzing the images because OCT produces many images for each patient. The classification of patients'
OCT pictures into eight categories—Choroidal Neovascularization (CNV), Diabetic Macular Edema
(DME), Drusen, Normal, Diabetic Retinopathy, Age-Related Macular, Central Serous, Macular hole—is
studied. To prevent patients from blindness, these diseases should be detected, diagnosed, and treated early.
The purpose of this research is to present a thorough analysis of how various retinal disorders are classified
using a variety of machine learning and deep learning techniques.
Keywords— Retinal diseases, OCT images, eight classes.
PAPER ID: CS418

Image Captioning for Information Generation
Gurvansh Vohra , Lakshay Gupta, Deepika Bansal, Bhoomi Gupta
[email protected]
ABSTRACT— In the AI applications for natural language definitions, image captioning is a field that is
expanding quickly. It attempts to capture meaningful interpretations of the interactions between the
acquired picture data from datasets and the sentence meanings. It combines Long Short-Term Memory type
RNN over phrases and CNN (Convolutional Neural Networks) picture reading techniques so that the image
can be understood by the system and output it in a common language. This study intermingles the vision
of the computer and natural language processing (NLP) to create machine that enhances visual information
such as photos by giving braille legible captions for those who are blind to better understand what is going
on around them.
Keywords- Computer Vision, Natural language, CNN, Image Captioning, LSTM.
PAPER ID: CS425

An Implementation of Blockchain Technology in Combination with IPFS for Crime Evidence
Management System
Shilpa.C, A.H., Shanthakumara
[email protected]
ABSTRACT— Crime is an unlawful act and is punished by the authority, in order to prove crime, evidence
is necessary. Evidence obtained from crime scene is crucial as it act as proof of crime, Digitalization of
36
ISBN: 979-8-3503-4821-7
evidences is need of hour. During the entire process of investigation heterogenous format of data is
generated and integrity of the sensitive data has to be maintained as sensitive data passes through the
various levels of intermediaries forming Chain of Evidences (CoE). Evidence needs to be tramper proof
and should be protected from any kind of alterations. In order to build strong system with immutability,
integrity, and legitimacy features blockchain technology is more suitable. The digital evidence can be
transferred in a transparent way between the parties involved without any central authority using blockchain
technology. This paper focuses on how blockchain based solutions can help in building a strong secure
system. The system is implemented using Ethereum platform to achieve integrity, immutability
transparency as well as tampering can be identified by any one at any time.
Keywords- Crime Evidence Management; chain of evidences (COE); Ethereum Blockchain; Smart
Contract; IPFS.
PAPER ID: CS426

Methods for Automatically Identifying and Diagnosing Mental Health Issues
Khadimul Islam Zim, Upinder kaur
[email protected]
ABSTRACT— People's reports of feeling content with their life have broad societal relevance since they
provide insight into what matters most to individuals and communities. To thrive, one must have access to
basic necessities like food, shelter, and gainful job. It is crucial for public policy to keep tabs on these
trends. However, many indicators that evaluate living conditions fail to capture what people believe and
feel about their lives, such as the quality of their relationships, positive emotions and resilience, the
realisation of their potential, or overall satisfaction with life (i.e. their "well-being"). Overall happiness and
how one feels on a daily basis, from sadness to elation, is what we mean when we talk about someone's
well-being. Emotional, psychological, and social well-being all contribute to what is known as "mental
health." How we feel, what we believe, and what we do are all impacted. How we respond to stress, interact
with others, and make decisions is also influenced by our personality. From childhood and youth through
maturity, maintaining good mental health is crucial. Focusing on the strengths, weaknesses, and potential
for improvement of mental health problem identification is central to this article. The goal of this article is
to provide a critical analysis of the literature on OSNs for the diagnosis of mental health issues. Here, we
look at the machine learning data source, the feature extraction strategy, and how well the classifier works.
In this paper we will review the previous done research on this topic and our paper will let everyone know
that which method give the best result. This study provide a road path for the future study on the mental
health.
Keywords- mental health, ML,SVM, CNN, KNN.
PAPER ID: CS427

Mental health detection and anticipation analysis using Deep Learning
Khadimul Islam Zim, Upinder kaur
[email protected]
ABSTRACT— There has been more research into how deep learning could help find, diagnose, and treat
mental health disorders because of how common they are, how important. When applied to the study of
human behaviour, deep learning techniques may reveal hitherto unexplored pathways, diagnosing mental
health issues, forecasting the course of diseases, and tailoring and improving treatment plans. Despite the
promising prospects for ML's use in mental health, this is still a nascent area of study, and there is a wide
range of complicated, intertwined problems that must be overcome in order to produce successful ML-
enabled applications that can be put into reality. This paper provides an overview of the existing deep
learning work on psycho-socially based mental health disorders in the computing and HCI literature and
37
ISBN: 979-8-3503-4821-7
aims to identify new paths for increasing growth in this crucial area. The investigation includes a
quantitative synthesis and a qualitative narrative assessment, which together revealed widespread patterns,
voids, and difficulties in this field. In this paper, we have covered the history of AI, ML, and DL models
for the diagnosis of mental health problems. All of these subtopics were selected because they are directly
related to the main subject. One of our key goals in writing this piece was to conduct research into the field
of mental health and identify the best deep learning model for early identification of mental health
problems.
Keywords- mental health, Deep Learning, Indian people mental health, SVM, CNN, KNN, Systematic
Review.
PAPER ID: CS428

Automated mechanism for improving precision irrigation using machine learning
Sasi kala Rani K, V.K.G Kalaiselvi, Hariharan Shanmugasundaram, Dwaraknath A
Swathi K, Tarrun R
[email protected]
ABSTRACT: To monitor and control the critical components of agricultural farming using the modernized
IOT Technology, AGRI - AUTOMATA emphasizes efficiency by increasing the yield by majorly looking
through humidity and temperature of the soil by even keeping a track of soil moisture content and sunlight.
Through this system, productivity is increased by enhancing the automated maintenance system. The
sensors are placed in equitable places in the farm, to sense and convey adequate information to the farmer
by comprising IoT technology through the help of required software to the user. This proposed system can
be used to improvise the productivity of crops and the supply of water and maintain brightness intensity by
even altering the temperature through the automated system with machine learning approach. Between the
incorporation of various sensors, the Node MCU microcontroller acts as the decision-making and
controlling device for the whole automated system.
Keywords: IoT technology; field Maintenance system; Node MCU microcontroller; Cloud Computing
PAPER: CS430
Voice Controlled Home Automation System using Google Assistants
Ajmeera Kiran, G. R. Sakthidharan, D Divya Priya, K.Prem Kumar, B.Uma Mahesh, K.Pavan
Kumar
[email protected]
Abstract— One of the most rapidly developing technologies that helps reduce the amount of work done
by humans is known as "home automation." People who are looking for luxury and sophisticated home
automation platforms are the target audience for some home automation systems. On the other hand, some
systems are designed to be user-friendly for people who have special needs, such as the elderly and the
disabled. Home automation and the use of applications that are controlled by voice are becoming more
common as a result of advancements in technologies such as Google Assistance and Amazon Alexa. Now,
we have developed our own version of this project, which is called "Voice Controlled Home Automation
Using Google Assistant." The goal of this paper is to design a home automation board that is both functional
and aesthetically pleasing, and that can be concealed within the AC power units that are mounted on our
walls. We are able to control all of the gadgets that are linked to the board by giving it a simple voice
command such as "Okay, Google, turn on the lights." This will cause the lights to turn on. Because of this
paper, we are able to control our household appliances online, manually, and using a timer.
Keywords- Smart Home, Artificial Intelligence, Home Automations, Amazon
38
ISBN: 979-8-3503-4821-7
PAPER: CS431
Remote Health Prediction System Using Machine Learning Algorithms
P Purushotham, Ajmeera Kiran, A.Punitha, D Divya Priya
[email protected]
Abstract— The "Remote health prediction System employing Machine Learning Algorithms" uses
machine learning in healthcare. Better diagnostic and treatment decisions require better medical facilities.
Machine learning helps clinicians analyse massive and complex medical data. This information helps
doctors treat patients. Thus, machine learning in healthcare improves patient happiness. This project
integrates healthcare machine learning functions. Illness prediction using machine learning predictive
algorithms can make healthcare smarter than diagnosis. Machine Learning Algorithms and libraries will
build this project. Due to its rich library and usability, Python is recommended. syntax. We used different
categorization approaches to train Machine Learning models and then chose the best model that avoided
overfitting and underfitting and had low variance and bias.
Keywords- Smart Health, Health Care, Machine Learning Algorithms, HER, Internet of Things
PAPER: CS432
Face Mask Detection Using Support Vector Machine
D Divya Priya, P Purushotham, R Subbulakshmi, Ajmeera Kiran, Arulananth T S, S.P. Kavya
[email protected]
Abstract— According to a report published by the World Health Organization (WHO), the only two
primary ways in which the COVID-19 virus can be spread are through breathing contaminated air and
through direct physical contact. Wearing a mask is one of the bare minimum preventative measures that
can be taken against the transmission of certain viral infections, including COVID-19. Therefore, it is much
more important to wear a mask properly in very crowded areas such as public parks, supermarkets, and
shopping malls. This is because the risk of spreading an infectious disease is significantly higher in these
settings. Face Mask Detection is a technique that we developed so that we could avoid problems like these.
It can determine whether or not a person is wearing a mask on their face. There have only been a handful
of scant studies done on face mask detection.
Keywords- Face mask Detection, Machine Learning, SVM, Covid-19, SARS
PAPER: CS433
Defect Exposure in Vegetables and Fruits using Machine Learning
S. Spandana, Ajmeera Kiran, M Sathiya, Kesavan Krishnan
[email protected]
Abstract— Defects to agricultural food items like fruits and vegetables, results in decrease of quality and
quantity of agricultural consuming items. This directly affects the financial source of agriculturists and the
human health. So, exposing to this kind of defects in fruits and vegetables at early stages of developing
products leads to minimization in loss of production and quality. The regular or most used day to day
approach for defect detection is continuously watching and observation of crop either by farmers or further
by some professionals. But , it is very time-consuming method. We have used image processing and
classification techniques in machine learning to identify and detect the diseases on agricultural products
like vegetables and fruits. The main goal of the system being proposed here is to monitor and detect diseases
in fruits and vegetables in their early stages to farmers for a healthy yield. Our System can deliver the most
accurate results in a very short time while consuming minimal computational resources.
Keywords- Vegetables Disease Prediction, Fruits Disease, Machine Learning Algorithms.
39
ISBN: 979-8-3503-4821-7
PAPER ID: CS438

Automated Answer Script Evaluation Using Deep Learning
Arepelly Shylesh, MD Abdul Raafeh, Shaik Mathin, Hariharan , Shanmugasundaram
[email protected]
ABSTRACT— Recent advances in deep learning are replacing conventional systems in many aspects of
life. Deep Learning provides a cost-effective and consistent alternative to human marking. An automated
script analyzer aids in overcoming the difficulties associated with manual evaluation. The increasing
competition in the education field made a heavy load on evaluators which is time-consuming. Finally,
irritation results in a difference in paper evaluation from student to student, which is not a reliable and
efficient way of evaluation. So to decrease manual power, increase technology for coping with long time-
consuming processes, and make equally distributed evaluations according to students writing and standard
of answers, we came up with the idea of automated scripting used for automated evaluation of scripts which
is an easy and effective way of correction. The system considers factors like grammatical errors, synonyms,
stop words, etc for analysis and gives scoring.
Keywords— Natural Language Processing (NLP); Deep Learning; Optical character recognition (OCR);
Convolutional Neural Network (CNN); Descriptive Answer Evaluation.
PAPER ID: CS439

A Comparative Study of Quantum Machine Learning Enhanced SVM and Classical SVM for Brain
Stroke Prediction
Nitya T. Singh , Aleena Swetapadma , Prasant Kumar Pattnaik
[email protected]
ABSTRACT — A stroke is a medical condition in which the blood vessels in the brain gets damaged
according to the WHO. It is also one of the most significant causes of death. Brain stroke can be prevented,
if necessary, measures are taken by the individuals who are at the higher probability of suffering from it.
In this work, a method has been proposed using quantum enhanced support vector machine for brain stroke
prediction. This prediction is done based on the medical history and the physical attributes of a body. A
model built which predict if a patient will suffer from brain stroke or not. In this work, quantum enhanced
support vector machine algorithms are also compared against the traditional SVM algorithms.
Keywords- Brain stroke, Support vector machine, Quantum computing, machine learning.
PAPER: CS440
Enhancing the Security for Healthcare Data using Blockchain Technology
Devi.T, S. Bangaru kamatchi, N.Deepa
[email protected]
Abstract— Cyber security is the safest way to protect the data from hackers and unauthorized users.
Healthcare technologies nowadays face lots of cyber security issues related with the security of the health
information and privacy of the data. Cyber security is one of the popularized ways to protect the data from
hackers and spammers. HealthCare is the field where the data is highly sensitive and the security for the
systems is low. Protecting the sensitive data is achieved by the blockchain method, which is similar to a
database but the difference is the data stored in the blockchain in blocks. The new blocks included are
connected to previous blocks from a chain like structure, it is very secure, each block stores data and also
the hash of previous blocks. So, data cannot be easily accessed or manipulated. The mechanism used in
blockchain for the security of the data consensus mechanism which contains the different methodologies
includes Proof of Work (PoW), Proof of Stake (PoS), Proof of Space and Proof of Authority. Enhanced
40
ISBN: 979-8-3503-4821-7
proof of stake is a combination of the PoS and DPoS used to increase the security of the system by
eliminating the 51% attack in blockchain and reduces the data theft threats and protects the medical records
of patients from hackers.
Keywords— Consensus, Proof of Stake, Blockchain,
PAPER: CS442
Sentimental Analysis Recognition in Customer review using Novel-CNN
N.Deepa, J.Sathya Priya, Devi.T
[email protected]
Abstract— Sentimental Emotion Recognition is the recognition of emotion which is detected by many
fields such as Artificial intelligence, Machine Learning and Deep Learning. The professionals need
sentimental analysis for business like social media monitoring, brand monitoring and customer feedback
which will help in business for improvising the product based on the emotion gathered from the customers.
Sentimental analysis is used in business to mine the data of customers about how they are feeling about the
product. Existing systems like Facebook, WhatsApp and twitter use sentimental emotions for sensing the
user’s emotions. In our proposed system we are using the Novel Convolution Neural Network (N-CNN)
system for sentimental recognition to make better understanding of customers which can be used to improve
the product quality. To enhance the accuracy of the proposed model Amazon review for product sales is
used which is available in kaggle. Using the proposed model the model in comparison with existing model
the preprocessed feature which are extracted from multiple neural networks are recognized. Feature based
on the selected customer feedback on constant steps of filtering, max pooling and necessary activation
function results are implemented and shown 98.3% of accuracy which is more reliable results than the other
Machine learning model.
Keywords— Artificial Intelligence, Deep Learning, Machine Learning, Novel -Convolution Neural
Network, Sentimental analysis.
PAPER: CS443
A Novel Two-way cross-tab Machine Learning Approach for predicting Life Insurance using
Bivariate Exploratory Analysis
N.Deepa, Udayakumar N, Devi.T
[email protected]
Abstract— Life Insurance prediction is the main objective of the research: to evaluate a person's life
insurance using a machine learning model. In day-to-day life human life becomes hectic and the life span
of everyone gets reduced due to pandemic situations, unavoidable accidents and historical impacts, etc.
Even though the security and saving beneficiaries are there in life insurance, risk factors are also associated
with it. Machine learning techniques propose a risk reduction avoidance and prediction of financial scams
to save customer's lives when individual customers claim their own experience in risk factors by sharing
their own credentials. The analysis is made by logistic regression based on the probability of categorical
data such as identity proof, Aadhar proof, PAN number, and so on as attribute value. Using the novel Two-
way cross-tab method the relationships based on attribute value matrix value are generated to find the
customer who has no mutual identifier to take the life insurance cash and summarize the prediction using
the bivariate exploratory flow of graph. By seeing the difference in the relationship, the threats and risk full
factors can be reduced in accuracy compared to existing machine learning models.
Keywords— Bivariate Exploratory, Financial Scams, Life Insurance, Machine Learning, Novel Two-way
cross-tab, Prediction.
41
ISBN: 979-8-3503-4821-7
PAPER:CS445
Shopify – Shop Augmentation and Recommendation System
Subhojeet Bhattacharjee, Nishant Singh, N. Prasath
[email protected]
Abstract— Modern times have brought about changes in lifestyle. People dislike visiting crowded markets
because they take too long and are uncomfortable. Thus, an advantage of online shopping is, it saves tons
of time. Online purchasing is a procedure where clients go directly to a vendor on the Internet to buy
products, services, etc. Customers can browse online merchants from the convenience of their home and
make purchases while seated at a computer. This research aims to introduce local people to the online
shopping application so that local merchants can profit from it and expand their reach and sales through
this fantastic platform. This research aims to give clients of physical stores the benefits of internet shopping.
It facilitates purchasing goods from any store using a smart device and the internet. In this way, the
customer will benefit from home delivery and online shopping services from his preferred retailer. The
solution is different in an umpteen number of ways. It is a direct exchange between the buyer and the
vendor. This system can be applied to any local business or to international brands with retail store chains
in the local areas.
Keywords- Encryption, Preference Matrix, Feature-based product analysis, Chatbot, AOT Compilation,
JPA, Scalability, Security
PAPER: CS447
Heart disease Prediction Using Machine Learning
Saurabh Bilgaiyan, Tajul Islam Ayon, Aliza Ahmed Khan, Fatema Tuj Johora, Masuma Parvin,
Mohammad Jahangir Alam
[email protected]
Abstract—Mainly related to the cardiovascular system, brain, kidney, and peripheral arteries, the disease
is called heart disease. Heart disease can have many causes, but high blood pressure and atherosclerosis are
the main ones. Additionally, structural and physiological changes in the heart with age are largely
responsible for heart disease, which can occur even in healthy individuals. They are not put to use or
employed in any way. If these data were investigated and examined, diseases may be predicted or perhaps
prevented. By using images of cancer cells to train a dataset, diseases like cancer may be identified and
their stage can be forecasted. Similarly, to that, factors like cholesterol, diabetes, heart rate, etc. can be used
to predict heart disease. It is difficult and dangerous to predict cardiac disorders. We noticed that sometimes
there are multiple approaches used to solve a problem. It varies depending on the circumstances. The fact
that most of the data are sparse or absent since they weren't recorded with the intention of analysis presents
another difficulty. With data from four hospitals in four distinct locations, we, therefore, set out to
determine which strategy would be best for forecasting the diseases. This study compares the effectiveness
of various data mining methods for predicting cardiac disease, including, Knearest neighbors, Random
Forest, and Multi-layer Perceptron, Logistic Regression. The effectiveness of prediction for each approach
utilized is reported after an analysis of the Data Mining methodologies. The outcome demonstrated that
heart problems can be predicted with greater than 97 percent accuracy.
Keywords- heart disease, k-nearest neighbors, Multilayer Perceptron, Random Forest, SVM, Logistic
Regression, Cardiovascular Diseases.
42
ISBN: 979-8-3503-4821-7
PAPER ID: CS449

Pitch Extraction and Notes Generation Implementation using Tensorflow
Vijay K, Krithiga P, Kavirakesh S
[email protected]
ABSTRACT— Great ideas are born in the least expected place. Musical ideas might just happen while
travelling, hiking or working or in any random place just not in the music studio. Every musician's desire
is to capture the flow of ideas in the moment to convert it into sheet music which can be passed on to their
instrumentalists.
Thus, it has become a research challenge to automatically detect tunes and recognise modes of humming
melody which is an important guarantee for humming music retrieval and automatic transcription. Deep
learning has played a vital role in extraction of tune features and utilising them to produce the instrumental
tones although with some limitations. The data is given and it is pre-processed and explored. Our project
is trained using a Deep learning algorithm, Neural Networks. The attention mechanism makes the network
focus on the humming pitches and gets the note sequence by detecting most relevant tune features directly.
Then, the mode of the note sequence is recognized according to the result of the tonic detection and
automated tones for the specified instrument are produced.
PAPER ID: CS451

RFID-Enabled Automated Billing System For Retail Establishments
Akhash S, Vijay K, M. Jaeyalakshmi, Adharsh Anand, Eashaan Manohar
[email protected]
ABSTRACT--
Abstract—This paper proposes an approach to equip shopping carts with IoT devices to enhance the user
experience and boost sales, thereby increasing store profits. The billing counter is usually the most crowded
spot in the store and gets even worse during the holidays. As a result, people get frustrated waiting in long
queues for an indefinite amount of time. They may even avoid visiting crowded shops, which results in a
considerable loss to the businesses. To overcome this problem, our proposed system makes the billing
process more accessible. It makes it easier for the customers to view the products, their quantity, and the
total amount to be paid. The shopping cart or trolley will be equipped with a small canister enclosing
NodeMCU and various sensors, such as an RFID reader, load sensor, and buzzer. The NodeMCU is placed
and connected with these sensors to eschew the common difficulties faced by businesses. Each product in
the store is equipped with an RFID sticker. When the products are dropped inside the cart, they get scanned
by the RFID reader, and all the information about the item is sent to the company’s server by the NodeMCU
present. The authentication process regarding the product is done, and the filtered information is sent to the
smart cart Mobile app. The scanned products can be viewed by the customers in the app. The app shows
the total amount, the items selected, and their quantity. This solution saves a considerable amount of time
and allows the customers to know the amount to be paid beforehand so that they can be prepared with cash
to pay at the billing counter, or they can opt to pay through the app’s payment page, which connects with
the store’s existing payment page.
Keywords: RFID, Billing, IoT, Shopping cart
PAPER ID: CS452

Private Cloud Storage using Raspberry PI via Virtual Network Computing - An Analysis
Vijay K, Sabarish Abishek W R, Sabarish V U, Sanjeev Krishnan R
[email protected]
ABSTRACT— The world is becoming increasingly data-driven. People today need more data to analyse
and gain insights. Storing and accessing large amounts of data becomes difficult, so our study will use the
43
ISBN: 979-8-3503-4821-7
Raspberry Pi to access all data from any devices connected to the system. Both direct and cloud-based
channels of communication between the Raspberry Pi and the device are possible. To that end, we're
conducting this research with the hope of one day being able to use VNC cloud to connect to Raspberry Pi
remotely and use the device that's attached to the Pi from anywhere in the world. When it comes to keeping
tabs on and managing your gadgets, remote control systems are invaluable. When it comes to remote
accessing graphical displays and sharing Android applications, VNC (Virtual Network Computing) is by
far the most popular tool. Known as a "thin client," VNC is a display protocol that works across multiple
operating systems. In this paper, we deploy a system designed for remote desktop access via mobile VNC.
Keywords- Raspberry Pi, Virtual Network Computing, Remote Access.
PAPER: CS453
Chest Disease Classification Using Convolutional Neural Networks
Ms. Kavitha S, Thaarani S, Amrit Preet Singh, Santhosh G
[email protected]
Abstract— Chest diseases pose major health risks to people globally. Early diagnosis of these conditions
enables early treatment, which can prevent death. The healthcare system benefits greatly from the use of
Convolutional Neural Networks (CNN), particularly when it comes to early disease prediction. Xrays serve
as one of the key factors that accurately classify disorders of the chest. The prediction of chest diseases
such Atelectasis, Cardiomegaly, Lung Consolidation, Edema, Pleural Thickening and Pneumothorax from
X-ray images is the objective of this project. The CNN Model is used to analyze the disease classification
and the results are explained in terms of accuracy. Preprocessing with images can enhance the model's
accuracy. For that, we used some image preprocessing techniques which include Histogram Equalization,
Bilateral Filter, Gaussian Blur and Contrast Limited Adaptive Histogram Equalization. These techniques
were used to remove the unwanted noise from the X -ray images and improve luminance of the images
which leads to produce more accurate decisions. The dataset consists of 1 csv file and an X-ray image
folder that contains six classes of disease and 1,120 X-rays. Convolutional neural networks (CNNs) are
described in the research as a tool for diagnosing disorders of the chest. The architecture of CNN is
presented, as well as its guiding principles. Among those preprocessing techniques, Contrast Limited
Adaptive Histogram Equalization technique gave more accuracy which is nearly 91.2 %. Results that
compare accuracy and network training time are shown.
Keywords: CNN,Multi class,Chest Disease,X-rays
PAPER ID: CS454

WOMEN’S SAFETY WRISTBAND USING IOT
Adharsh Anand, Vijay K, M. Jaeyalakshmi, Akhash S, Akshita A
[email protected]
ABSTRACT Nowadays, it is essential to safeguard the safety of women in public places. Despite the
introduction and usage of several new technologies and preventative measures, it continues to be a global
issue. Formation of effective rules and regulations by the Government to prevent and control the crime rate
against women are turning out to be ineffective. The government cannot be blamed as the safety of women
is the responsibility of the people. Law enforcement agencies often keep track of criminal records, which
are not always readily available to the general public. Although certain wearable technology and
smartphone apps have been created to combat violence against women, they don't exactly guarantee a
woman's protection in times of need. Our main goal is to safeguard them before the criminal incident
happens. Thus, we created a mobile application and a wristband to monitor the activities of the user with
44
ISBN: 979-8-3503-4821-7
the help of the Geographic Information System (GIS) and to investigate, track, and stop the crime by
alarming volunteers close to the incident cluster. This GIS by locating the crime hotspot and maps out its
tendencies. The proposed system analyzes crime hotspots and stops it before it happens in the near future
using data collected from a mobile app and a wristband. The wristband is very beneficial for women who
feel insecure to travel alone. Any unfortunate situations can be avoided by using this wristband. Since the
safety of the women is the top priority of our system, services our provided 24/7 and throughout the year.
The data gathered from the website and mobile app is further collected and processed for analysis purposes.
The users of this mobile app will be able to know the crime rate of the place they are about to visit.
Keywords— Women, Safety, Wristband, Crime, Investigate
PAPER: CS455
Design and Implementation of Alzheimer's Disease Detection using cGAN and CNN
Dr.R.Ravikumar, Dr.T.Thilagaraj, Hareesh Raj R, Gokula Kannan G, N Sasipriyaa, Abishek A
[email protected]
ABSTRACT: The importance of diagnosing Alzheimer's disease (AD) and Mild Cognitive Impairment
(MCI), its prodromal form, for patient care, has made it a new study focus because of AD's high prevalence
and huge social impact on health. Existing research have used machine learning approaches to diagnose
AD early by looking at numerous biomarkers. However, in the bulk of current research, the diagnosis
performance is hampered by the inability to efficiently extract high-level brain features that indicate the
primary abnormalities of anatomical brain regions connected to AD. In order to diagnose AD brain pictures,
the suggested model combines a deep convolutional neural network (CNN) and a conditional generative
adversarial network (cGAN).The dataset is preprocessed to create a clear representation of the brain's grey
matter, and the cGAN is utilised to create more training instances for the model to improve accuracy.
According to the limitations of the discriminator, the generator in cGAN-ADD learns to include the illness
category feedback from the classifier into the 2Dbrain slice image reconstruction process for image
improvement in the training phase. In the prediction phase, the stacked CNN layers extract high-level brain
properties from categoryenhanced 2D-brain slice pictures. The classifier gets the collected brain features
in order to output the posterior probability of disease conditions (Normal, AD, Mild and Very Mild).
Tensorflow is used to train the model, and Keras is used to evaluate classification accuracy. The model is
trained and tested on ADNI dataset.
Keywords: Alzheimer Disease Detection, Mild Cognitive Impairment, Convolutional Neural Network,
Conditional Generative Adversarial Network, Deep Learning
PAPER: CS456
A Deep Learning Approach Based on CT Images for an Automatic Detection of Polycystic Kidney
Disease
S.K.Aruna, Devi.T, N.Deepa
[email protected]
Abstract— Deep Learning is a subdivision of machine learning; deep learning is mainly separated into
multiple neural network layers. These layers allow it to learn large amounts of the data and try to mimic
the human brain. By adding more hidden layers it can help to improvise the efficiency. Deep Learning
techniques are utilized to identify diabetes mellitus, vascular dementia early diagnosis and mammography
nodule ultrasonographic recognition. Pathology and Radiology will be invested in the future for new
advancement in deep learning. Kidney irregularities is one of the most common diseases among people of
all ages. A global shortage of urologists developed a deep learning model to diagnose kidney problems
with this study. The kidney area is surrounded by a thick wall. It is possible to have multiple cysts with
high moisture concentrations that cause renal diseases. The units from zero -twenty indicate different types
45
ISBN: 979-8-3503-4821-7
of renal diseases. Tissues that can be grouped together form a solid lamp known as tumor. Tumors can
cause direct damage to glands, spinal cells and other organs. Urolithiasis is another name for renal stone
disease where a large amount of solid materials enter into the digestive tract. In a research paper, CNN is
trained to learn how the cysts diagnosis should be done. Tree Convolution neural deep learning techniques
VGG19 are used to detect kidney abnormalities. To train and evaluate a dataset of 4000 abdomen computed
tomography with classifier and segmentation is created. The increased categorization and accuracy rate
obtained to display the efficacy of kidney disorders is 98%.
Keywords— Convolutional Neural Network(CNN), Computed Tomography, Deep Learning, Cysts Kidney
Diseases
PAPER ID: CS457

Recognition of Tamil handwritten characters using Scrabble GAN
N SASIPRIYAA, DR.E.GOTHAI, DR.P.NATESAN, G.MADHESAN , E. MADHUMITHA ,
K.V.MITHUN
[email protected]
ABSTRACT— Recognition of hand- or machine-printed documents is now a crucial component of
applications. A commonly used method for converting a printed or handwritten file into its equivalent text
format is Optical Character Recognition (OCR). Handwriting recognition is a major research area in pattern
recognition and attracts extensive research in OCR. In Southern part of India, Tamil is the primary language
with the longest continuous literary heritage. OCR is used to extract characters from scanned digital input
images and convert them into formats that can be edited by machines. It is quite challenging to recognize
handwritten Tamil characters due to variations in font size, orientation angle and writing styles, bigger
category sets, and misunderstanding over the similarities between handwritten characters. In this study, we
suggest a deep learning-based technique for handwritten Tamil character recognition. Text documents
printed on paper require time-consuming character correction and reprinting. Here, a model incorporating
scrabble GAN with CNN is used to increase the dataset. New training instances are created from the old
instances using Scrabble Generative Adversarial and Networks augmentation to improve the accuracy of
detecting the Tamil handwritten characters. The primary goal of the research work is to expand the
handwritten characters and recognize them. Hyperparameters such as kernel size, activation function,
padding, step size and layer count are adjusted to enhance CNN-based handwriting recognition. In order to
distinguish between characters, CNN evaluates their differences in shapes and features..Dataset consists of
nearly 6,000 images and the 80% of the images are used as training images and 20 % of the images used
as testing images and obtained the accuracy of 96.23%.
Keywords- CNN, Scrabble GAN, GAN, Deep learning, Tamil handwritten character recognition.
PAPER ID: CS461

SMART DUSTBIN USING NODEMCU
A Adesh Sai, Vijay K, M. Jaeyalakshmi, Anjana Deivasigamani
[email protected]
ABSTRACT— In today’s world, urbanization has grown significantly. There is an increase in the
production of garbage on the same segment. Waste management has been a crucial subject to take into
account. This document enables us to successfully establish the desired results. The smart bin in this
instance is built on a microcontroller-based NodeMCU board that is interfaced with an ultrasonic sensor.
Cleaning is the first requirement of a wise and healthy lifestyle, and cleaning calls for a trash can. A
community will only properly dispose of its trash if dustbins are also mandated to be filled. At the top of
the trash can is an ultrasonic sensor that will gauge the stature of the trash can. The threshold stature is set
at a specific height. Due to the real-time management of smart Dustbins, it will prevent dustbins from
46
ISBN: 979-8-3503-4821-7
overflowing along roadsides and in neighborhoods. The waste can be rapidly controlled to its efficient level
once those clever containers are used on a large scale with the help of replacing the conventional containers
because it prevents pointless lumping of rubbish at the roadside. Long-term issues could also result from
the foul smell of such rotten wastes that are left untreated for a long time due to authorities' neglect and the
public's irresponsibility. By breeding in a dirty environment, pests and mosquitoes can be a nuisance. This
might potentially inspire terrible diseases. This project aims to make environmental maintenance simple
and efficient. It also strives to create a seamless complement to green spaces.
Keywords- NODEMCU; IFTTT; webhooks; Ultrasonic Sensor.
PAPER ID: CS468

IMPROVED NETWORK SECURITY BY UTILIZING CLOUDY AUTOMATED
WATERMARKING TECHNIQUES
Dr. J .Viji Gripsy, K R Kanchana
[email protected]
ABSTRACT— A Mobile Ad-Hoc Network (MANET) is made up of a number of wireless mobile nodes
that lack any predetermined infrastructure or topology. This is independent of any centralized access point
or management. Due to these traits, it suffers from significant network overhead and energy-related
problems. Several routing methods have been created and developed in the literature to reduce routing
overhead and to increase MANET's packet delivery ratio A Secure Route Discovery-Adhoc On-demand
Distance Vector (SRD-AODV) protocol with effective authentication via the Modified Elliptic Curve
Diffie-Hellman Algorithm has been proposed (MECDHA). It is also required to validate node energy and.
So, in this article, we introduced Cloudy Automated Watermarking Techniques (CAWT). In this techniques
is used to create a cloud. This cloud based on automated watermarking techniques. The technology of
automated watermarking is quite secure. A key inserts watermarks into nodes. Only someone who is aware
of the keys involved in embedding will be able to remove the watermark nodes. Finally, this reduces
network routing overhead while improving security and efficiency in energy minimization and data
delivery.
Keywords — MANET, Routing, Cloud based Routing protocol
PAPER: CS471
The Early Detection of Dementia Disease Using Machine Learning Approach
Md Shariar Kabir, Jakia Khanom, Md. Atik Bhuiyan, Zerin Nasrin Tumpa, SK. Fazlee Rabby,
Saurabh Bilgaiyan
[email protected]
Abstract— Early dementia detection is a crucial but challenging task in Bangladesh. Often, dementia is
not recognized until it is too late to receive effective care. This results in part from a lack of knowledge
about the illness and its signs and symptoms. Recent improvements in machine learning algorithms,
however, may change this. In a recent study, we developed a model that can identify early dementia in
Bangladesh using machine learning algorithms. This research paper proposed an efficient machine
learning-based approach for early detection of dementia disease A dataset of 199 people with dementia and
175 healthy controls was used to develop the model. In 96% of the cases, the algorithm correctly identified
dementia. This is a significant accomplishment that could revolutionize Bangladesh's dementia detection
process. For patients to get the care they require, early dementia detection is essential. This study offers a
proof-of-concept for the use of machine learning in dementia early detection & The results of this study
suggest that machine learning models can be used as a powerful tool for early detection of dementia.
Index Terms—Dementia, Machine Learning, Prediction, Accuracy
47
ISBN: 979-8-3503-4821-7
PAPER: CS473
KNOWLEDGE BASED PERCEPTIONS OF CHATBOTS IN HEALTHCARE
Nagaraj P, Muneeswaran V, Manoj G, Mohammed Akhil Sk, Akshay Kumar N, Venkata Teja G
[email protected]
Abstract— This order to gauge a happy life, it is essential to have a healthy body. However, it might be
difficult to get a doctor's appointment for any type of health condition. The objective is to use AI to build
a health chatbot that can identify illnesses and provide basic information about them. before calling a
physician. It is possible to reduce healthcare costs and improve access to medical information by using a
medical chatbot. Computer programs known as chatbots interact with users orally. The chatbot stores the
data to identify the sentence keywords, select a question, and respond to the query. Each sentence's score
is determined from the given input sentence, and other similar sentences will be found.
Keywords—: Chatbot, Healthcare, Artificial Intelligence, Perceptions, Knowledge Discovery
PAPER ID: CS474

YOLO BASED LICENSE PLATE DETECTION OF TRIPLE RIDERS AND VIOLATORS
Nandhakumar R G, Dr.N.Krishnamoorthy, Pranesh V R, Dr. K. Nirmala Devi, Dr.S.Shanthi,
Nikhalyaa S
[email protected]
ABSTRACT— Road safety is of prime importance as road accidents are among the biggest causes of
deaths in the country. Road Accidents are majorly due to violators and lawbreakers of road safety rules like
not wearing helmets, triple riding etc. Even though there are many smart systems to monitor these
violations, it is complicated to keep a track of the data and view it efficiently anywhere. Also monitoring
the permit details of other state vehicles is hard. A method was proposed to address this issue. based on
Deep learning and Optical Character Recognition (OCR). Here we detect the riders not wearing helmet and
triple riders using object detection by comparing YOLOv2 and YOLOv3. This simplifies the task of traffic
police who can’t continuously and efficiently monitor all the violators. We gathered our data from three
different sources. Kaggle was used to collect the data set. A number of items can be recognized by the real-
time object recognition system YOLO in a single frame. It recognizes object more precisely and faster than
other recognition systems. It can predict up to 9000 classes and even unseen classes. It divides the image
into m*m matrix and each grid detects objects within itself. The trained model is tested using a test dataset
set after training in order to measure the performance. For Number plate detection we use OCR based
character recognition. The recognized license plate images of violators are captured and stored in a database
which is then sent to the concerned department. The stored data can be easily accessed using the developed
mobile app. Metrics considered for evaluation is Accuracy. The average is utilized to assess object detection
models like R-CNN and YOLO. The average calculates a score by comparing the detected box to the
ground-truth bounding box. The better the score, the more precise the model's detections are. When
compared to YOLOv2, YOLOv3 exhibits strong performance in terms of accuracy with 95.5%.
Keywords- Deep learning; Object detection; Convolutional Neural Network; YOLOv2; YOLOv3
PAPER: CS475
FAKE NEWS DETECTION USING MACHINE LEARNING APPROACH
P. Santhiya, S . Kavitha, T. Aravindh, S. Archana, Ashwanth V Praveen
[email protected]
Abstract: - The emergence of the World Wide Web and the quick uptake of social media platforms (like
Facebook, Instagram and Whatsapp) have made it possible for information to be disseminated in ways that
have never previously been seen in the history of humanity. Online hiring has changed the hiring pattern.
In particular, putting job adverts on corporate websites and career portals includes looking for a sizable
48
ISBN: 979-8-3503-4821-7
pool of qualified candidates worldwide. Unfortunately, it has become yet another forum for scammers,
which threatens applicants' privacy and harms the reputation of companies. The topic of detecting recruiting
fraud and scams is addressed in this case study. Three machine learning models are used in the construction
of an effective recruitment fraud detection model, which includes a number of significant organizational,
job description, and kind of remuneration features. The proposed system uses different Three Machine
learning approaches like Support Vector Machine, Random Forest, Naive Bayes Classifier to determine
whether a job posting is genuine or false. We extracted features from data using two techniques: Term
Frequency-Inverse Document Frequency (TF-IDF) and Bag-of-Words (BoW). As a result, three models
achieve decent results. An ensemble model is created by training three independent machine learning
models using different segments of samples and then using a simple majority vote of the three models to
determine the final predictions. As a result, three models achieve decent results and an accuracy of more
than 98.18% have been achieved using Random Forest Model.
Keywords - Machine Learning, Supervised Learning Models, Ensemble Model, Evaluation Metrics
PAPER ID: CS476

FAAP -A Farmer’s Amiable APplication to enhance Agriculture using Mobile Technology
Gunavathie M A , Dr. K. Sasi Kala Rani, M. Manikandan, G. Renugadevi
[email protected]
ABSTRACT— The farmers are in need of a pandemic app that can make their livelihood serene. The goal
is to make the farmer's work even easier by analyzing the minerals in the sand. The farmer is even tutored
with methods of crop cultivation that are suggested and crop species that are rare and the one that produces
high yield is suggested for cultivation. In this paper, we presented “The farmer-friendly app to increase
their livelihood” which focuses on the overall perspective. The app gets the geographical location as the
input. It then analyses the input to produce various attributes that are indeed essential for the overall gross
increase in the agricultural economy. Machine learning is the technology used to train the app according to
various heterogeneous data that are available. With this app, the farmer not only knows the resource,
minerals, and crop variety but also is backed up with the remedies that would eventually prevent them from
facing extensive suffering.
Keywords- Farming, agriculture, crop, yield, mobile app
PAPER ID: CS477

Cervical Cancer Detection Using the Ensemble Method in Machine Learning
Roopa C , Dharaneesh KT , Ganeshkumar S ,Hari Vignesh KG
[email protected]
ABSTRACT— In the recent years most common gynecologic malignancies worldwide is still cervical
cancer. Early screening is the most efficient method of reducing the worldwide burden of cervical cancer
because it is a disease that is highly preventable. However, due to limited knowledge and access in
developing nations, the vulnerable patient must go to medical facilities and undergo very costly operations.
The majority of people cannot afford to be checked out frequently. An ensemble strategy is offered to
assess the possibility of cervical cancer in this paper. The ensemble techniques like bagging including
random forest, boosting including Adaptive Boosting and Extreme Gradient Boosting, Light gradient
boosting machine, and other effective algorithms such as Decision tree and Support Vector Machine are
showcased. This method is used to overcome the challenges that previous cervical cancer studies
encountered. To improve performance, a method for data rectification is proposed. Since the ensemble
strategy performs better than single classifier, this proposed approach is more scalable and useful than
others.
Keywords— Cervical cancer, Adaptive Boosting, Extreme Gradient Boosting, Support Vector Machine.
49
ISBN: 979-8-3503-4821-7
PAPER ID: CS478

Suicidal Thoughts of Recent Generations- A Review
Dr .Shanthi. N, C.Roopa, A.Aadhishri, S.K.Surya Prasath
[email protected]
ABSTRACT— Individuals are now more immersed in their virtual lives than ever owing to the
proliferation of social networking sites, and the massive amount of data that people post online makes it a
home for researchers performing their studies. People prefer modern technology and use online social
media as a means of communication while they are having suicidal thoughts. The social issue of suicide
has received a lot of attention recently. Social media data analysis using machine learning offers a viable
method for identifying long-term contextual factors that increase a person's risk of having suicidal thoughts
and actions. This paper describes the review of suicidal risks and factors by considering various papers and
the influence of social media like twitter data of past few years. As text descriptions and user language are
analyzed, a wide range of data that can be used as a key warning system for suicidal identification is
revealed. Several features are retrieved to identify tweets displaying suicidal thoughts, and a collection of
features is provided for training the model over the dataset using various machine learning algorithms.
Keywords— social media, suicidal thoughts, machine learning
PAPER ID: CS479

Emotion Prediction through Facial Recognition Using Machine Learning: A Survey
Kripansh Singh, Naman Goel, Bhoomi Gupta, Deepika Bansal
ABSTRACT— An application created by integrating Flutter and machine learning is one that predicts or
manages stress. Its sole and primary function is to identify the stress and pressure that today's population
is experiencing and recommending them ways to deal with it. Humans are constantly activated by a variety
of unnoticed environmental stimuli. Sometimes these stressors are undetectable. However, the machines
are capable of handling this well, allowing us to create a programme that can make decisions instantly.
Today, a lot of scientists and researchers use the information gathered from the internet to help them make
the right choices. In order to determine if a person is under stress or not, we are once again using image
recognition techniques and camera data in this application. This software takes a picture with the camera,
processes it, then decides what to do with it and displays it on the screen. There will be a different section
in application where multiple methods to deal with stress will be suggested.
Keywords- Stress Prediction, Facial Recognition, Machine Learning, Stress Relief
PAPER ID: CS481

SMART DIAGNOSTIC SYSTEM FOR EARLY DETECTION AND PREDICTION OF
POLYCYSTIC OVARY SYNDROME
Harish K P
Dhivyanchali M.N, Dr.K.Nirmala Devi , Dr.N.Krishnamoor, Dhana Sree R, Dharanidharan R
[email protected]
ABSTRACT— PCOS are the serious condition which affects female ovaries during their reproductive age
of 15 to 45 . This disease affects 5 to 10% of reproductive-age females. Although it is difficult to fully
resolve this issue ,the PCOS affected women can be mitigated through proper exercise, by taking proper
nutritious diet and maintaining the healthy BMI. Until they take a pregnancy test, the majority of women
50
ISBN: 979-8-3503-4821-7
are unaware of the disease. The clinical dataset has the 541 instances and 45 attributes of unbalanced classes
of 0 and 1 (no and yes) which has 364 instances of 0 (no) class and 177 instances of 1 (yes) class .
Preprocessing is done for the unbalanced dataset by filling the null values and changing the datatype of all
attributes to numeric datatypes. The unbalanced dataset is balanced by the balancing techniques of SMOTE
and Random Over Sampling. Comparing the both balanced techniques through the accuracy the random
oversampling gives the best .The supervised learning algorithms are Decision tree, KNN ,Random Forest
,AdaBoost, Logistic regression ,Gradient boosting, cat boosting, XGBoosting , Linear SVM, Radial SVM,
Linear discriminant analysis and Quadratic discriminant analysis are used.The supervised learning
algorithms are trained and tested by splitting the dataset to 70% for training and 30% for testing . The
ensemble stacking techniques are used by implementing the all models at the cross validation of 10.The
xgboost gives the accuracy of 96% for the balanced dataset.
PAPER ID: CS482

An approach to Detect melanoma skin cancer Using fast ai CNN Models
Md Shazzad Mia, Sumaiya Mustari Mim, Mohammad Jahangir Alam, Md Razib, Mahamudullah,
Saurabh Bilgaiyan
[email protected]
ABSTRACT— Skin cancer, which can be lethal, is essentially the improper proliferation of skin tissues.
It has recently developed into one of the most dangerous sorts of additional malignancies in humans. Early
detection may help the patient endure. Skin cancer is notoriously difficult to detect. Currently, computer
vision performs incredibly well when used to diagnose medical images. Along with technological
development and the rapid rise in computer accessibility, several machine learning algorithms and deep
learning algorithms have been developed for the interpretation of medical images, especially images of
skin lesions. According to our paper, there are five fast-ai CNN pre-trained models with various image pre-
processing techniques that enhance the classification capability of skin lesions and make them more precise
than other existing models. The HAM10000 dataset's benign and malignant cancer lesions are distinguished
by utilizing a number of pre-processing methods. The experimental findings showed that the suggested
model improved its accuracy to 97% in both training and testing.
Keywords- Skin cancer, CNN, fast-ai, Melanoma, HAM10000.
PAPER: CS484
Prediction Of Typhoid Using Machine Learning and ANN Prior To Clinical Test
Md. Atik Bhuiyan, Sharaf Shahariare Rad, Fatema Tuj Johora, Abdullah Islam, Md Ismail Hossain,
Ms. Aliza Ahmed Khan
[email protected]
Abstract—One of the most prevalent illnesses, typhoid causes a large number of fatalities each year,
primarily in Africa. A quick and accurate diagnosis is essential in the medical sector. Selfmedication,
delayed diagnosis, a lack of medical expertise, and inadequate healthcare facilities all contribute to the high
incidence of typhoid fever mortality. Machine learning as well as deep learning has worked wonders for
extrapolative analysis in the health industry, and as a result, more health industries are utilizing machine
learning techniques. This is the earliest evaluation where a typhoid fever prediction model is being
developed which predicts prior to a clinical trial. In this paper, deep learning and machine learning have
been employed to develop the model. Ten algorithms have been utilized here, and the XGBoost classifier
is the best performer with 97.87% accuracy.
Keywords-typhoid fever; machine learning; artificial neural network; XGboost.
51
ISBN: 979-8-3503-4821-7
PAPER ID: CS485

Text Detection and Language Identification in Natural Scene Images using YOLOv5
G.R. Sreekanth, R.C. Suganthe, R.S. Latha, R. Logesh, R. Rajadevi, V.V. Jagadeeswaran
A. Maheshvar
[email protected]
ABSTRACT— Deep learning has immensely evolved ever since digital era. Deep learning also includes
feature extraction as a facet. Text snipping from a picture is a difficult task since the image comprises text
in a variety of sizes, styles, orientations, alignments, low contrast, noise, and with a complicated backdrop
structure. Transformation of an image into different perspective for feature identification is the first step
towards text recognition. Scene texts provide rich contextual information that can be applied to several
types of vision-based applications, hence over the last few years we have witnessed an increase in interest
in the detection and recognition of scene texts. In order to address the issue of language detection from
multilingual scene text photos, a deep learning-based solution is suggested in this paper. In this study, the
underlying model of a Convolutional neural network is employed to detect objects in real-time with high
accuracy. This study employs a single neural network “you only look once" known as YOLO, since it offers
predictions with just a single forward propagation trip through the neural network to evaluate the full image.
We used COCO ‘Common Objects in Context’ dataset which is a large-scale object detection,
segmentation, and captioning dataset. To evaluate the image YOLO divides the image into smaller parts
and forecasts boundary areas and probabilities for every part. The predicted probability weighs these region
proposals. It then provides identified objects after non-max linear suppression. We used F1-score which
combines accuracy and recall into a single metric by computing their harmonic means.
PAPER ID: CS488

THRESHOLD BASED FEATURE SELECTION FOR ANEMIA PREDICTION
S. RAMYA, G.Brindha, E.B.B.Boomika, K.Aishwarya
[email protected]
ABSTRACT– The imf›lementation of intelligent software to identify and predict diseases at the earliest is
growing importance in many Jields. India iS one of the develof›ing countries that contributes ttiost to the
major global health problem of’ anaemia. The Machine Learning system %r Anew ia prediction could
predict the presence oJ anemia Jor the f›eoples in emeigenc y department. Many new algot-ithm
are achieving a great in prediction.An algorithm's accuracy, precision, recall, Fl-Score, and supf›ort ore
all systemnticall y a,s,sessed. There nre otily n rclativcly small number of techniques for predicting
anaemia, including decision tt-ees, i’andom J!orests, naive Bayes, etc. Thus, anaemia is predicted using a
variety of methods, including XGBoost, LightGBM, La.s.so, Rigde, and Ela.stic-net. For foreca.st.s, these
worked well. By randomly increa.sing rind duf›licating minority class .saitiplcs, SMOTE seeks to balance
the distrihution oj classes. It is dotie to improve accuracy by choosing/’eatures. To determine the strongly
associatcd points thnt are the threshold in feature.selection, Pear.son correlation i.s u.red.Here, LightGBM
performed well and gives the dataset a higher accuracy rating. In order to determine correctne.ss, our
J›roposed approach will u.se the Skilcnrslan dataset. To forecast anaemia, the dataset has 15,000 row.s atid
more than 20 classes. This dataset includes o[15,300 p‹itients'full blood count test i esults Jroin between
2013 and 2018. The outcome demon.strates that the Light GBM algorithm surpasses other algorithms in
terms o[ accuracy.
Keywords– machine learning, prediction, accuracy, algorithms, and SMOTE.
52
ISBN: 979-8-3503-4821-7
PAPER: CS489
RECURRENT GASTRIC CANCER PREDICTION USING RANDOMIZED SEARCH CV
OPTIMIZER
Vishnu M K, Vishal Rupak V R, Vedhapriyaa S, Ms.M.Sangeetha M, Dr.R.Manjuladevi,
Ms.C.Sagana
[email protected]
Abstract—Gastric cancer has the second-highest mortality rate and the fourth-highest incidence of
malignant tumors worldwide. Early stomach cancer has a fair prognosis, but the symptoms are unusual and
the signs are not immediately apparent. The goal of the project was to use patient clinicopathological data
and machine learning techniques to increase the precision of a prediction model for the emergence of gastric
cancer. The current study proposes the first four parameters. By comparing the performance of Gradient
Boosted Decision Tree, Random Forest, Decision Tree, and Gradient Boosting Machine, Logistic
Regression, it was shown that machine learning approaches can be used to estimate the repetition of patients
with malignant after such operation. The parameters influencing postsurgical repetition of stomach cancer
have been proposed in the present study. In this case, Random Forest is enhanced using the Random Search
Cross Validation optimizer to increase the model's performance in predicting cancer. Random Search Cross
Validation performs a random search over parameters, sampling each setting from a range of potential
parameter values. The dataset contains 2012 patient records as well as 51 characteristics.
PAPER ID: CS490

Data Security and Authentication in Modern Applications and Cloud Computing
M. Hemanth Kumar, Virat Anand, Ankit Seth, Achyuta Kumar Mohapatra , T.Vignesh
AV Praveen Krishna
[email protected]
ABSTRACT— Data security is the technique of preventing digital data from being accessed by
unauthorized parties, being corrupted, or being stolen at any point in its lifespan. It is a notion that covers
all facets of information security, including administrative and access controls, logical security of software
programmers, and physical security of hardware and storage devices. Organizational policies and
procedures are also included. Robust data security measures, when executed correctly, guard against insider
threats and human mistake, which continue to be among the primary causes of data breaches in the modern
day, while also safeguarding an organization's information assets from cybercriminal activity.
Implementing tools and technology that improve the organization's visibility into where its crucial data is
located and how it is utilized is a key component of data security. In a perfect world, these technologies
would be able to automate reporting, apply safeguards like encryption, data masking, and redaction of
sensitive information, and apply protections like encryption, data masking, and compliance with regulatory
standards. Personal information, such as credit card numbers, Social Security numbers, driver's license
numbers, and medical histories, as well as business information, including customer lists and source code,
are frequently exposed in data breaches. Identity theft or a breach of industry or governmental compliance
regulations may occur from a data breach, and the offending corporation may be subject to penalties, legal
action, reputational damage, or even lose its ability to conduct business. There are several ways that data
breaches can be countered in the modern day and there are techniques which are yet to be implemented in
this field. As we can see the importance of preventing data breaches, we conducted a research on
implementing security mechanisms in applications. Our study explores the various ways that data breaches
could occur in contemporary online applications and offers solutions to the issues raised.
Keywords- Data Security, Authentication Systems, Encryption, Threat detection, Cloud systems.
53
ISBN: 979-8-3503-4821-7
PAPER ID: CS495

A Novel Method for Pneumothorax Diagnosis and Segmentation Using Deep Convolution Neural
Network
S.Shanthi, K. Nirmaladevi, N. Mathiaras, AS. Krithika Tharani, A. Monisha, R. Kuppuchamy,
N.Krishnamoorthy
[email protected]
ABSTRACT— A pneumothorax is the accumulation of air in the pleural space between the chest and
lungs. Air that is in the lungs during breathing can leak out of the lungs and become trapped between the
chest and lungs in the thoracic cavity, causing chest pain, difficulty breathing, and ultimately the complete
collapse of the taut lungs. pneumothorax. Therefore, early diagnosis and treatment should be started. Neural
networks have proven effective in image classification tasks, and many researchers believe that training
neural network models can detect radiologist-diagnosed chest abnormalities such as pneumonia and
pneumothorax. is proving. Moreover, we utilize various techniques, including Stochastic Weight
Averaging (SWA), aggressive data augmentation, and learning rate schedule. We use the chest x-ray dataset
provided by the 2019 SIIM-ACR In this study, a ResNet-34 neural network is trained on a data set of chest
x-rays containing over 10,675 images to create a model that predicts the presence of pneumothorax from
chest x-rays.our suggested system receives a Dice Similarity Coefficient (DSC) of 0.8410
Keywords- Pneumothorax, Image Classification, Image Segmentation, ResNet34, U-Net, Transfer
Learning
PAPER : CS502
Twitter Sentiment Analysis using Collaborative Multi Layer Perceptron(CMLP) Classifier
Dr.R.Manjula Devi , Dr.P.Keerthika,, Dr.P.Suresh,, Ms. M. Sangeetha ,Ms.C.Sagana, Ms.S.Savitha,
Mr.K.Devendran, B.Nithiesh
[email protected]
ABSTRACT In the last decade, Twitter is facing a major challenge in identifying the sentiments of the
tweets. These tweets are useful in understanding the opinion of different end users about a variety of topics.
Sentiment Analysis is more important to analyse the text data in the dataset. But the tweets consist of non-
useful characters which make sentiment analysis difficult.
With multi-layer perceptron classifier, classification for new tweets as either positive or negative is done.
This can be done by using a range of features like tokenisation, lemmatisation, stemming on the words used
in the tweets, the sentiment of the words, and the user’s profile. Feature extraction is made to clean the
words/phrases containing non-useful characters like URL, Numeric alphabets, punctuations etc. It
improves the accuracy for the large amount of data with high efficiency.
Keywords- Twitter Analysis, Multi layer perceptron, sentiment analysis
PAPER : CS504
PREDICTION OF PARKINSON’SDISEASEUSING MACHINE LEARNING BASED ON VOCAL
FREQUENCY
N.P.Saravanan, P.Dhanush, P.Deepika, P.Dhanvarsini
[email protected]
ABSTRACT— Parkinson's disease (PD) is a neurological disorder that affects most people after
Alzheimer's. The ageing neurodegenerative condition leads to PD that reduces dopamine levels in the brain
are a defining feature. Tremor, stiffness, bradykinesia, and postural instability are the core characteristics
of PD. PD patient’s(PWP) quality of life is impacted by both motor and non-motor symptoms, which may
also have an indirect impact on family and caregivers. The patient's quality of life can be improved and
54
ISBN: 979-8-3503-4821-7
maintained with an early PD diagnosis. There is currently no cure, however there are therapies to control
the illness, such as dopaminergic medications.
Speech problems associated with PD are categorized under the term hypokinetic dysarthria for the
diagnosis of PD. The traditional approach is based on their clinical history and also with their physical
examination. With the help of automatic analysis tools, practitioners may diagnose patients, monitor their
progress, and conduct regular, economical assessments that are objective. This study uses a machine
learning approach to conduct a pilot experiment to identify the existence of dysarthria in speech and gauge
its severity. With the help of SVM Algorithm and Hyper Tunning displays 93% accuracy on KAGGLE
database test samples.
Keywords- Dysarthria detection · Machine learning · SVM.
PAPER : CS505
Machine Learning-Based Crop Recommendations for Precision Farming to Maximize Crop Yields
C.Sagana, K.Devendran, P.Mithun, M.Sangeetha, T.Kavin, S.Savitha, K.Kavinsri
[email protected]
ABSTRACT — Technology is now playing a critical role in many industries to overcome challenges and
achieve stronger and maximum results. The farming industry in India significantly affects the country's
economy. The agricultural sector continues to employ half of the nation's workforce. The natural conditions
of the environment have a big impact on the agriculture industry, which creates a number of difficulties in
actual farming operations. The majority of agricultural practices in the nation still use outdated methods,
and technological advancement is slow. Appropriate technology can be applied to this industry to maximize
yield and minimize challenges. Rather than considering the parameters like soil conditions, sustainability,
etc., farmers frequently sow the crop based on its market value and potential financial profits. This might
have unfavorable effects on both soil quality and farmers. If used properly, modern techniques like machine
learning techniques have the potential to revolutionize these fields. This paper will demonstrate how to use
such technologies effectively to offer farmers the best crop recommendation support.
Keywords- crop recommendation; machine learning; soil nutrients; rainfall
PAPER : CS506
SALES PREDICTION FOR FOOD PRODUCTS USING MACHINE LEARNING
Mr.N.P.Saravanan, R Muthupriyanka , K C Nandha Kishore , S Ishwarya
[email protected]
ABSTRACT - Machine learning contain many algorithms to predict outcomes more accurately without
having to be explicitly coded. By using machine learning algorithms, the prediction is done with the help
of the parametric evaluation and conclusion matrix. Nowadays the competition among the supermarkets is
increasing and the customer expectations are increasing. In this paper the overall sales is identified and
with more accuracy and the other extra classification of the attributes are also predicted, then the prediction
of the specific product also identified and location wise sales, supermarket sales and other categories for
the more analysis of the sales is also predicted and in the overall sales which range the sales is high also
predicted and the several algorithms like KNN algorithm, Naïve Bayes algorithm, and Some algorithms
are used for the prediction.
55
ISBN: 979-8-3503-4821-7
PAPER: CS508
A Perlustration on the optimistic prognosis of chronic renal failure using Evolutionary Machine
Learning and Deep Learning techniques
Vishnupriya B, Agalya M, Abitha R, Gugapriya M, Harshithaa A S
[email protected]
Abstract —Millions of individuals worldwide die from chronic kidney disease each year as a result of bad
lifestyle choices and inherited factors. A kind of kidney illness known as chronic kidney disease (CKD)
sees a slow decline in kidney function over months or years. Initially, there is typically no symptoms. 's
machine learning algorithm is a vital tool for medical practitioners to reliably identify illnesses at an early.
In the framework, multiple machine learning algorithms are used to analyze renal disorders, and the best
ones are chosen to determine the severity of CKD patients. For machine learning, it utilized a dataset on
chronic renal illness from the UCI machine learning repository, which contained 25 functions. The
Decision Tree, Support Vendor Machine, K Nearest Neighbors, Logistic Regression, and Naïve Bayes
machine learning algorithms are employed in this work. Ensemble Learning has been applied based on
boosting algorithm. The results were validated using four scoring techniques: Accuracy, Precision, Recall,
and F1 measurement. Before categorization, the data were scaled to normalize the range of independent
variables and pre-processed to manage missing values. X-ray or Magnetic Resonance Imaging (MRI)
pictures of the kidneys are used as 28x28 pixel data for deep learning. Picture filtering and image
segmentation are used as part of the image pre-processing approach. A local ACED and a better Sobel on-
the-edge detection technique are then used. Finally, a model is developed to forecast chronic kidney disease
based on the result.
Keywords— Chronic Disorder, Machine Learning, Deep Learning, Decision Tree, Support Vendor
Machine (SVM), Logistic Regression, Naive Bayes, Ensemble learning, Edge detection, Improved Sobel,
Local ACED.
PAPER: CS510
Whale Species Classifier Using Dorsal Fin And Lateral Body Characteristic
Kunal Pamu, Dr Allam Mohan
[email protected]
Abstract— Most of the time whale classification is a huge problem because of the low availability of the
training data. The researchers working on whale species identification have stated that manual recognition
of the whales is not that easy because the whales do not surface very often, the only time they surface to
the water is when they have to breathe through their blowhole. This is one of the many opportunities of
manual recognition used by the researchers to capture an image of the whales. The dataset used in this
paper is comprised of the images of whales containing the characteristic of ‘Dorsal Fin and Lateral Body’.
In this Paper the classification of ‘Whale Species’ using the images of dorsal fins and lateral body is
performed using a pre-trained model Mobile Net v2 based on CNN and ANN. The preliminary
classification model made in this paper achieved the test accuracy of 87.15%.
Keywords— Whale Species Classification, Deep Learning, Artificial Neural Network, Image
Classification, Convolutional Neural Network, Multiclass Classification.
56
ISBN: 979-8-3503-4821-7
PAPER: CS511
Heart Disease Prediction-based on Conventional KNN and Tuned-Hyper Parameters of KNN: An
Insight
Sanjeeva Polepaka, R. P. Ram Kumar, Srujan Katukam, Sai Vinay Potluri, Sunell Dutt Abburi,
Mohith Peddineni, Nagaraj Islavath, Muralidhar Reddy Anumandla
Abstract— The proposed research work aims to predict the Heart Disease-based on conventional KNN
and tuned-hyper parameters of KNN for enhanced performance. Initially, the conventional KNN Model is
adopted to determine and predict the heart diseases with respect to the dataset collected from Kaggle
Domain. Later, the hyper-parameters are tuned for determining the prediction performance on the same
dataset. The various parameters identified for tuning the performance includes, the parameter grid, score
and cross validation. Experimental observations depict that the hyper-tuned KNNs algorithm has enhanced
performance than the conventional KNN model in terms of prediction accuracy.
Keywords-KNN, Supervised Machine Learning, Hyperparameters, Precision, Recall, F1 – Score, Accuracy
PAPER: CS512
Alive Human Detection in War Fields and Earthquakes using PIR Sensor and Monitoring the
Heartbeat of the Victims
Kanchi Mishra, Ismath Tabassum, Sanjeeva Polepaka, R. P. Ram Kumar, Amal K
Abstract—During war fields and natural disasters like earthquakes, rescue crews struggle to find victims
due to people being buried under debris, stuck in inaccessible locations, or unseen by rescue helicopters.
In addition, even after locating the victims, the rescue team lacks information regarding their health. The
project's objective is to identify answers for such circumstances. It comprises a robot equipped with sensors
that identify victims and then determine their heartbeat rate. The human body emits infrared waves that are
detected by a Passive Infrared (PIR) sensor, which detects the living human body. After detecting a human,
the robot attempts to determine the victim's health status by monitoring the heartbeat rate. Additionally, the
prototype is linked to Bluetooth, which transmits all captured data to the Mobile Application. The rescue
teams analyze the received data to make further decisions. The project provides an efficient solution to
problems encountered in war zones and natural disasters such as earthquakes, along with information
regarding the victim's health status. In addition to performing activities such as object detection in the path
using ultrasonic sensors, the robot is capable of detecting alive humans using sensors. Keywords-PIR
Sensor, Heartbeat Sensor, Bluetooth, Ultrasonic Sensor, War fields, Natural Disasters, Mobile Application.
PAPER: CS514
Performance Analysis of Blockchain Technology using Inter Planetary File System (IPFS) database
in comparison with Block chain database
Meenu Jain, Dr. Manisha Jailia, Dr Manisha Agarwal
[email protected]
Abstract— Blockchain, as the name suggest is a long chain of blocks which are connected using the hashed
of the data which is stored in the previous blocks. In present scenerio schools,colleges and universities are
being forced towards the creation, maintainence, retrieval, safety and privacy of educational records.
Challenge is not only to store and maintain the data but also ensures fast access of data to all the
stakeholders. Blockchain based transcripts save the time and money and smart contarcts help in the
authentication.But there are several limitations of blockchain database , When sharing blockchain data from
a server to multiple clients, it is highly likely that multiple clients will need the same data. They are working
on the same case and may be investigating the same wallet. Blockchain and IPFS based transcripts save the
time and money and smart contarcts help in the authentication. Transcripts are safeguarded by blockchain
in the event of a server outage or natural disaster. This work analyse the perfoemance of IPFSdatabase that
57
ISBN: 979-8-3503-4821-7
uses content based addressing and cryptoghrapic hashes with other database.In this paper IPFS is compared
with blockchain database. IPFS is hypermedia protocol that is both safe and easy to use. It uses the content
based adressing with the merger of cryptographic hashes to work upon the distributed file sdystem .This
paper compares IPFS with Blockchain database upon the following factors like Size of files shared by
nodes , Degree of linking through number of peers linked, Bandwidth used.
Keywords-Blockchain, Interplanatary File System
PAPER ID : CS515
SYSTEMATIC HARMFUL SIGNS DETECTION FOR WOMEN’S SAFETY USING NEURAL

NETWORKS
U.Yaswanth Reddy, M.Sai Nikhil, Srithar S, P.S Sai Krishna
[email protected]
ABSTRACT— Nowadays, there was an increase in the extent of disruptive things and offensive things
performed toward people, especially women. Due to this, safety has been given preeminent importance and
foremost significance [1]. Public locations like purchasing centers, malls, banks, parks, theaters, and
numerous others are more and more places are geared with CCTV cameras to make sure safety and security
for Women. Constant examination of the CCTV footage by people was an unachievable task for them to
locate the danger in it. While it needs a lot of personal and steady interest to find whether the captured
things contain suspicious or not suspicious activities in them. This is one of the disadvantages of seeing the
danger in the images or videos of things that are captured in them by human beings and their people want
to identify the object and which components of the footage include the odd recreation which recognizes the
earlier identification of whether every motion is suspicious or not. Therefore, to reduce the time and
workforce of humans, we are introducing the use of a ML algorithm recognized as the Darknet 53
algorithm. Its purpose is to find out the symptoms of the weapons in the actual period, which filters out
irregularities from regular or common signs. which channels out inconsistencies from normal signs. From
that point, the place where we can locate or find an alert for the situation of savagery, showing the dubious
things or activities at the precise time of event period on women.
Keywords- CCTV Cameras, Suspicious, Images, Darknet, Weapons, Women.
PAPER: CS516
Classification of Satellite Agriculture Imagery to Identify Crop Type Using Deep Learning
Techniques - A Review
N.Thamaraikannan, Dr.S. Manju
[email protected]
Abstract— DL is a subdivision of machine learning, which is a subset of AI.Over the past two decades,
deep learning has gained a great deal of attention. Deep learning requires large quantities of data and
produces accurate results. Agriculture is a new field of study; yield prediction is particularly critical in
agriculture and is largely dependent on soil and climate elements including moisture, temperature, and
climate. The potential of planetary exploration is becoming extremely relevant in crop monitoring. As the
incidence with which satellites acquire remote sensing images increases. Today, Rapid changes in
environmental circumstances, on the other hand, have made it impossible for the farming community to
continue doing so. Crop prediction in agriculture is a complicated process one of main challenges that
researchers in this field face is the lack of cultural labeled data that is synchronized throughout spatial and
temporal. Evolutionary computation models have piqued the interest of many agricultural experts due to
58
ISBN: 979-8-3503-4821-7
their assurance in machine vision. Effectiveness in hyperspectral imagery (HSI) recognition for a variety
of earth environmental monitoring, including Agro - based measuring and land surface identification
Limited frequency components associated with process entity attributes are blocked by extract features.
Retrieved and embedded high bonding ghastly spatial highlights addressing the progressive design of target
entity attributes, providing the model with improved classifier achievement and reduce computational
Complexity.
Keywords-Crop classification, Hyperspectral images, Deep learning, Agriculture. Convolutional Neural
Networks, Recurrent Neural Networks.
PAPER : CS519
System for Detecting Intrusions using Raspberry PI
Dennise Mathew, Kirubasri G, Vijay K Eugene Berna I ,Sowmia K R
[email protected]
ABSTRACT — Home invasions and thefts have gotten far too frequent in recent years. However, it is
uncommon for people to be familiar with different systems, such as intrusion detectors, surveillance
systems, and so on. The average person finds it challenging to maintain his home secure as a result. CCTV
cameras are expensive in surveillance since they employ computers. It sets aside an excessive amount of
space for ongoing recording and needs personnel to find irregularities. Our goal is to develop a
comprehensive home security system that will help homeowners protect their properties from such
disasters. The Raspberry Pi4-based Home Security System project aims to let someone secure their home
against burglary and theft in one go. However, the Raspberry Pi system is significantly less expensive, has
greater resolution, and has capabilities that use less power than the current system. Here, infrared (IR)
sensors are used as straightforward yet effective triggers for human presence. For tiny confidential data
area surveillance, this solution is appropriate. i.e., a home office, a bank restroom, etc. The image is taken
by the camera and momentarily saved in the Raspberry Pi module whenever motion is detected by the PIR
sensor within the room. Applications built on the Internet of Things can be used remotely to monitor
activities and receive alerts when motion is detected.
Keywords- Raspberry Pi system, PIR sensor, Infrared sensors
PAPER: CS520
Multi-class Classification of Insects using Deep Neural Networks
Santhiya M, Priyadharshini A, Agshalal Sheeba J, Dr.S.Karpagavalli
Abstract— Insects are crucial to the functioning of nature. There are more than a million described species
of living beings in the modern world. Since the majority of today's farmers and agriculturalists are newer
generations of people, identifying and classifying insects is essential. The classification of insects is a
difficult undertaking in the agricultural industry. In the proposed work, multi-class classification of insects
using a Convolutional Neural Network architecture, VGG19 had been carried out. In the taxonomic
classification of insects, 5 insects fall within insecta class which include butterfly, dragonfly, grasshopper,
ladybird, and mosquito data had been collected to train, test, and validate the convolutional neural network,
The performance of the model had been analyzed using different parameters and presented.
Keywords — Insect classification, Taxonomic classification, multi-class classification, Deep Neural
Network, Data Augmentation, Convolutional neural network
59
ISBN: 979-8-3503-4821-7
PAPER : CS521
Copy-Move Image Forgery Detection Using Machine Learning
Aditya Semwal, Anurag Kumar, Ms.D.Saveetha
[email protected]
ABSTRACT — In the era of social media, where information can be shared with a single click, it is
important to segregate data into real and fake. Since the majority of the population uses the internet it is
easy to spread misinformation. Digital information in the form of photos is one of the primary modes of
communication. Unfortunately it has become increasingly easy to manipulate them. In our paper, we survey
the recent proposed methods of CMFD(Copy-move forgery detection), and explained the whole CMFD
process done. Particularly, we characterize a CMFD workflow for feature extraction and the matching
process using key points and descriptors.
Keywords- Copymove, forgery detection, SIFT, Agglomerative clustering
PAPER: CS522
Multi-Object Detection and Tracking Using Machine Learning
Ayush Sahay, Keerti Vardhan Singh R, Dr. J. Godwin Ponsam
[email protected]
Abstract— In today's information-based economy, data serves as a replacement for oil. There have been
shifts in how quickly and accurately benchmarks are measured as a result of efficient data. Industry
buzzwords Computer Vision (CV) and Artificial Intelligence (AI) perform the data processing, making the
improvement observable (AI). Two technologies have allowed for previously impossible endeavours, such
as the identification and tracking of objects for traffic surveillance systems. The need for an effective
algorithm to unearth concealed elements in images grows in tandem with the number of such features.
Single-object detection in the urban vehicle dataset is handled by a Convolutional Neural Network (CNN)
model, while multi-object detection in the COCO and KITTI datasets is handled by YOLOv3. Metrics are
used to evaluate and chart the performance of the models (mAP). On traffic surveillance video, we use
YOLOv3 and SORT to follow objects as they move between frames. This study argues that cutting-edge
networks like DarkNet deserve to be treated as special cases. We see effective detection and tracking on a
dataset of urban vehicles. The algorithms produce very reliable identifications that can be used in real time
for traffic applications.
PAPER ID:CS525
An Real-Time Object Detection Method for the Visually Impaired Using Machine Learning
Saravanan Alagarsamy, Ch. Sandya Niharika, T. Dhiliphan Rajkumar, D. Usha rani, K. P. L.
Syamala, K. Balaji
[email protected]
ABSTRACT — Vision, one of the five fundamental human senses, is crucial for defining how people
perceive the objects around them. Visual impairments affect more than 200 million people worldwide,
severely limiting their ability to perform numerous activities of daily living. Thus, it is essential for blind
people to understand their surroundings and the objects they are interacting with. In this work, we created
a tool that helps blind persons recognize diverse items in their environment by utilizing the YOLO V3
algorithm combined with R-CNN. This comprises a variety of approaches to develop an app that not only
instantly recognizes different objects in the visually impaired person's environment but also guides them
using audio output. A convolutional neural network (CNN) called YOLO (You Only Look at Once)
60
ISBN: 979-8-3503-4821-7
recognizes objects in real time. This suggested method is more effective and accurate than other algorithms
for recognizing things, according to research results, and it produces results for object detection that are
extremely similar in real time. It is crucial for persons who are blind or visually impaired to be able to
reliably and effectively detect and recognize objects in order to navigate both common and unfamiliar
situations safely, become stronger, and become more independent.
Keywords— Convolutional Neural Network, Computer vision, Hyper Text Markup Language
PAPER ID:CS526
Detection of Animals for Road Safety Using Deep Learning Method

Saravanan Alagarsamy, Kurapati Pavan Chand , R. Rajasekar, Kumbhagiri Nikhil, S. Korrapati
Mohan Krishna, Kota Venkata Pavan Teja
[email protected]
ABSTRACT — Wild animals must be monitored effectively and consistently in their natural
environments. This project creates an algorithm to find animals in outdoors. The effort of manually
recognizing all the various species might be challenging due to their diversity. This method categorizes
animals based on their photos so that we can more effectively watch them. Animal tracking, animal
accidents, and theft can all be avoided with the use of animal identification and classification. Deep learning
algorithms can be used effectively to identify wild people. This process helps to save wild animals and also
to prevent vehicles from accidents while crossing the road. The deep learning method alerts the users
traveling in vehicles by locating the animals nearby to the vehicles passing the road.
Keywords— Animal Detection, Classification, Deep Learning Algorithms, Template Matching Algorithm.
PAPER: CS529
Automatic image Captioning using CNN and LSTM
G.R.Sreekanth, R.S.Latha, S.Aruna, M. Indrajith,V, J. Kalanjiya Vishnu, S. Manoj
[email protected]
Abstract— The use of intelligent software to detect and categorize things, such as CNN and LSTM, is a
technology that is becoming increasingly important in many industries, including social media. Deep
convolution networks excel at object detection and classification. The current study created a deep
architecture of convolutional neural networks to improve the precision of object identification. There is no
special technique for accurately classifying object images using CNN at this time. Items are classified using
a variety of well-known CNN and LSTM-based deep neural networks in this case. These networks
performed well when it came to object classification, but locating objects in an image required the use of
sliding windows. It took a long time to process each window just to process one image. With one neural
network, this model learns the actual picture, divides it into target areas, and then accurately indicates
coordinates and likelihood of occurrence for each area. The metrics from the results computed across the
test set's 5000 photos object-relation Transformer that we proposed beats transformer of standard type in
across all criteria. The scores varied by an amount that was technically significant for BLEU-1, BLEU-4,
CIDEr-D, ROUGE-L, Count (p = 0.05) and Relation and the large increases in CIDEr-D score and Relation
support our hypothesis that adding the geometric attention layer will aid model for to be making most
appropriate object connections. Furthermore, the Count subcategory of SPICE has improved significantly.
61
ISBN: 979-8-3503-4821-7
PAPER ID:CS30
Animal Detection, Classification, Deep Learning Algorithms, Template Matching Algorithm.

S. Saritha, G. Abel Thangaraja
[email protected]
ABSTRACT— Agriculture is an essential, important sector in the wide-reaching context. Farming helps
to satisfy the basic need of food for every living being. Agriculture is considered the broadest economic
sector. The crop yield is a significant part of food security and improves the drastic manner by human
population. The quality and quantity of the yield touch the high rate of production. Farmers require timely
advice to predict crop productivity. The strategic analysis also helps to increase crop production to meet
the growing food demand. The forecasting of crop yield is a process of forecasting crop yield by using
historical data. Machine learning provides a revolution in the agricultural field by changing the income
scenario and growing an optimum crop. Many researchers carried out their research to deal with forecasting
crop yield. In this way, accurate prediction of crop yield was improved. But, failed to reduce the crop yield
prediction time and the accuracy level was not enhanced by existing methods.
Keywords — Agriculture, economic sector, strategic analysis, crop yield prediction, machine learning.
PAPER: CS32
Effective Model for Improving Symptoms-Based Disease Prediction by BiMM - BERT Algorithm
V.Jayasudha, N.Deepa, Devi.T
[email protected]
Abstract— Nowadays, symptoms-based disease prediction is more helpful for predicting the correct type
of disease. These predictions are more helpful to the medical industry and doctors; they can reduce the time
taken to check up and accurately predict the disease type. Combining the Bidirectional Encoder
Representations from Transformers (BERT), Natural Language Processing (NLP), Hidden Markov Model
(HMM) and Bi-directional Maximal Matching algorithm (BiMM), a unique word classification algorithm,
BERT-BiMM, is presented. It can use quick word segmentation to match sub-dictionaries. The knowledge
hidden in many research topics can be found using Natural Language Processing methods. The knowledge
hidden in many research topics can be found using NLP methods. The remaining text is joined by the
remaining specific words BiMM post text segmentation; the remaining text is connected by the remaining
single words using Hidden Markov Model. We compared the real-world clinical text. We took some
symptoms of the disease. HMM, BiMM, achieves maximum accuracy. BERT is used to connect the
remaining text with these methods, and high accuracy in medical text segmentations will be obtained if this
model outperforms provincial models by achieving strict scores of 91.29% on the BiMM model.
Keywords— Natural Language Processing, Bidirectional Encoder Representations from Transformers,
Hidden Markov Model, Hidden Markov Model, Text Classification, Word Segmentation, Disease
Prediction.
PAPER ID:CS536
Comparative Analysis of different Approaches to the Music Recommendation System

Nagaraj P, Muneeswaran. V , Mohith Kumar B, Rama Krishna Rao K, Mothukuri Siva Nagaraju,
Mandadi venakata hemanth kumar
[email protected]
ABSTRACT— Digital music and music streaming services have led to a significant expansion in the
variety of music available. For anyone, sorting through this music would be impossible. Music
62
ISBN: 979-8-3503-4821-7
recommendation systems that use genre, artist, instrument, and user evaluations to automatically suggest
music greatly reduce the amount of manual effort required. Although commercial use of music
recommendation systems is widespread, there is no perfect system that can offer the customer the greatest
music recommendations with the least amount of human effort. This research examined various
recommendation systems currently in use, including content-based, collaborative, emotion-based, and other
approaches. In addition to discussing the various recommendation methods, this research also explored the
advantages and disadvantages of each approach. Finally, an overview of a potential music recommendation
system that addresses many of the challenges faced by current hybrid recommendation systems was
provided.
Keywords: Recommendation system, music recommendation system, content-based filtering, collaborative
filtering, hybrid filtering, session-based recommendations.
PAPER ID:CS538
Recommendation system, music recommendation system, content-based filtering, collaborative

filtering, hybrid filtering, session-based recommendations.
A.Akhil, A.Priyanka, Y.Ajay, Srithar S, B.Vandana, Prem kumar Ch
[email protected]
ABSTRACT— In this technological era automation is playing a vital role in making things faster. Face
Recognition is one of the examples of automation. Facial Recognition is categorized as a biometric method
alongside fingerprint scan and iris scan. Back in the days facial recognition is only limited for safety and
security purposes but nowadays it is widely used in various industries like Banking, Healthcare, and many
more. One such industry is the educational industry. Various universities around the globe have already
implemented biometrics through facial recognition not only for the staff but also for recording the student’s
attendance. Currently available Facial Recognition systems are able only able to track one person at an
instance. The constructed system must track the individual attendance while he/she being recorded in a live
video or while they were in a group photo using machine learning techniques. The system must analyze
each individual in a group using the image provided and mark the individual’s basic information like name,
class, section, etc. In general, the system is trained with each individual’s 180-degree facial view with the
help of convolutional neural networks. And to improve accuracy the system must be trained using deep
learning and deploying with a huge and distinct dataset. For achieving this we have conducted a study
comparing five well-established algorithms to know which algorithm is reliable
Keywords- Extraction, Facial Recognition, Neural Networks, Deep Learning, Classification.
PAPER: CS540
Prediction of Rental Demands using different Machine Learning Algorithms
Ravikumar S, Vijay K, S. Pavithra, Dr S Prithi, Sriram Kannan
[email protected]
Abstract— A Bike rental expectation conjectures the interest for bikes rentals in reliance of climate
conditions like the temperature and calendric data. In bike rental frameworks, it is vital that the managers
have to realize the number of bikes to be placed in each and every station. This include empowers them in
or-ganizing the necessary number of bikes at each station and choose whether a specific station needs to
have additional number of bikes or it is not re-quired. In this venture we will foresee the no of bikes required
for explicit climate and climatic conditions.
63
ISBN: 979-8-3503-4821-7
Keywords— Decision tree algorithm, Machine Learning, Random forest algorithm, Regression, Rental
Demands.
PAPER ID:CS542
DETECTION OF PERSONAL ATTACKS IN MOBBING ON SOCIAL MEDIA USING DEEP

LEARNING TECHNIQUES
Dr. P.A.Selvaraj, Dr.M.Jagadeesan ,Dr.T.M.Saravanan, Y.Shaheela, Suraj Kumar Pandey
M.Sowmithra
[email protected]
ABSTRACT— Mobbing is a serious problem on the Internet that affects both grown-ups and teenagers.
miscalculations similar as despair and self-murder followed. There's a growing demand for content
regulation on social media platforms. The following work builds a model grounded on detecting
cyberbullying from textual data using natural language processing and machine literacy using two different
types of cyberbullying data, detest speech on Twitter and commentary grounded on particular attacks on
Wikipedia forums. Three characteristic birth ways and four groups are considered to determine the most
effective system. The model provides further than 90 percent content for tweet data and further than 80
percent content for Wikipedia data.
Keywords Mobbing; CNN ; Regression; Deep Learning; Social Networks.
PAPER ID:CS543
Exploring the Potential of Blockchain Technology for Securing and Managing Student Records
Ms. Meenu Garg, Dr. Neha Agrawal, Vivan Bawra
[email protected]
ABSTRACT – Keeping track of students' data is tedious in big countries like India. Different generations
used different technologies and adopted various methodologies in organizing and storing this data. This
data should be kept secure for diverse purposes so that it can be utilized in the future. Thus, colleges and
universities spend a handsome amount of money managing and organizing this data. Blockchain is the
new technology of the future that will certainly conquer the world. Adopting blockchain for the storage of
data can be done so that the data remains on a platform that is accessible by everyone yet secure. The
project associated with this research paper is an ideal approach to storing and managing the students’ data
in a particular university or institution.
Keywords – Blockchain, Ethereum, Solidity, Smart Contracts, Web3.js, Ganache, Transactions.
PAPER: CS544
GLAUCOMA PREDICTION USING SUPERVISED LEARNING
R.C.Suganthe, M.Geetha, R.S.Latha, S.Preethi, M.Rishika, D.Sandhiya
[email protected]
Abstract— Glaucoma is an eye disorder when the optic nerve suffers damage due to an increase in
intraocular fluid pressure. When there is too much fluid present or the channel in the eye through which
the fluid normally drains is blocked, pressure builds up. People blind due to glaucoma and estimated that
4.5 million people globally affected by glaucoma. Among all global blindness, blindness happen by
glaucoma is slightly more than 12%. It is the second most frequent reason for blindness in the globe. Percent
of affected people that are not even aware of having glaucoma, may up to 50% in developed countries and
64
ISBN: 979-8-3503-4821-7
may rise to 90%in undeveloped parts of the world. So, detection of glaucoma is a crucial role in life
Glaucoma can be detected using increased cup to disc ratio where optic cup is a cup in the center of the
optic disc. The optic disk is the circular area inside the eye where the optic nerve connects to the retina.
The Glaucoma prediction can be done with Machine Learning, Deep Learning Techniques. The proposed
work proposes Convolutional Neural Network algorithm which gives a great classification performance
when compared to other algorithms such as Support Vector Machine (SVM) and Random Forest.
Keywords-component; Glaucoma, Random Forest, Convolutional Neural Network,Deep Learning, Support
Vector Machine ,Machine Learning, Classification, Segmentation, Preprocessing.
PAPER ID:CS545
Automated Stock Price Prediction Using LSTM-ANN
Nagaraj P, Muneeswaran V, T. V. Dastagiri Reddy, B.Venu Gopal Reddy, T.Ganesh Reddy
P.Suresh
[email protected]
ABSTRACT— It is well known that the stock market is erratic, dynamic, and nonlinear. Correct stock
price forecasting is extraordinarily difficult due to more than one (Large and small) factor, such as politics,
world economic circumstances, surprising actions, and business’s economic performance. But, there's also
this has the potential. There is a statistic toward discovering patterns. This gave an upward jab to the idea
of Trading using pre-programmed, automated buying and selling methods known as algorithmic trading.
In the generation of massive information, bottomless knowledge for expecting inventory market expenses
and tendencies has become still more famous than before. We accumulated 10 years of records from Yahoo
Finance and anticipate charge trends of inventory markets, a deep learning-based model and comprehensive
customization of characteristic engineering are presented. The planned answer stands complete as it
consists of preparing the inventory for forecasting inventory market price trends, a market dataset, several
function engineering methodologies, and a customized deep learning-based device combined. The model
achieves a typically high level of stock accuracy in market estimates. We also build a streamlet application
for users to easily access that they can search a stock ticker and our model predicts the previous 100 days’
values and make a detailed overview of the 101st day.
Keywords- Long Short-Term Memory, Asset Management Companies, Moving Average.
PAPER ID:CS546
Comparing and Describing the Characteristics of Swarm Optimization in Bio-inspired Algorithm

G. Likitha, Madala Phanindra, D. Rithvik Venkata Sai, Srithar S, Aravinth S S, V. Sravya Chandana
[email protected]
ABSTRACT— The majority of medical data research includes bio-inspired algorithms to identify the best
answers, mainly swarm optimizations techniques. Up to this point, there are nine swarm optimization
approaches derived through publications. All these bio-inspired algorithms are based on animal behavior,
lifestyles, and environments. To directly locate the best procedures, we must overcome various challenges
to choose an optimal answer for medical research. Even if we can identify some algorithms as the best
methodologies, for various research projects, we choose various algorithms as the major ones depending
on the qualities that they require. As a result, we examine all the optimization strategies' characteristics in
detail in our work. The readers can determine the optimization strategy they want to apply in their
application by reading the qualities they want to incorporate in their project since many optimization
65
ISBN: 979-8-3503-4821-7
strategies cannot be anticipated as the best approaches based on a single criterion. As a result, we covered
the characteristics of swarm optimization strategies, took five main characteristics of bio-inspired
algorithms, and mentioned the best suitable algorithm for each of them.
Keywords- Speed, Hunt, Search, Adaptability, Swarm
PAPER: CS548
IOTSDL: INTERNET OF THINGS SECURITY FOR DEEP LEARNING TECHNIQUES-A
RESEARCH PERSPECTIVES
Selvam R, Velliangiri S
[email protected]
Abstract: The Internet of Things (IoT) is an important component of information technology. New trend
technology is being developed during the Industñal 4.0 Revolution. IoT connects physical things from a
variety of industries, including smart homes, wearable technology, vehicular ad hoc networks (VANETs),
Healthcare and Smart Cities. The number of IoT applications has dramatically increased recently, and it is
predicted that by 2030, there will be 131 billion linked devices, an increase of 13% each year on average.
Security is a crucial issue in today's IoT industry. In order to detect Internet of Things attacks and identify
new types of intrusion to access a more secure network, using deep learning techniques in various models
is a useful tool. This helps to overcome the challenges associated with securing IoT devices. The need for
developing an attack-identification and classification system for intrusion detection. We provide a study
on the distinction between legitimate and malicious actions in order to distinguish abnormalities and
intrusions, as well as Network traffic analysis to find new threats. This paper provides broad reviews of
deep learning for Internet of Things Seurity. The Major contributions are helpful for researchers and
academicians for further research in the direction of Internet of Things Security. This reviews studies by
assessing their effectiveness using two kinds of fresh real-time traffic information. (i.e. Bot-IoT datasets
and CSE-CIC-IDS2022). We evaluate ‘accuracy rate’, ‘detection rate’ in many system.
Keywords - Deep Learning (DL); Internet of Things (IoT); Intrusion Detection System (IDS); CSE-CIC-
IDS2022; Bot-IoT
PAPER ID:CS550
Prediction of water quality parameters by IoT and machine learning

Srinivas Kolli, M. Ranjani, P. Kavitha, D. Arul Pon Daniel ,Awadhesh Chandramauli
[email protected]
ABSTRACT- Increases in domestical applications and other applications drive the need for better clean
Water Management in India. What makes up the "chemical, physical, and biological" quality of freshwater
There are three stages in the conventional method of monitoring water quality: sample, analysis, and
inquiry. The scientists do these by hand. This method needs to be foolproof and predict a water supply's
quality. They’re Using the development of wireless sensor technologies, has also been some investigation
into the possibility of monitoring water quality with wireless sensors put in water and communicating with
farmers through brief messages regarding water. There has also been a study into the possibility of
employing machine learning algorithms to analyze water quality. In light of recent advances in Machine
Communication, which allow for devices to communicate with one another and smart ly analyze the data
they collect. We have developed an "smart water quality monitoring system by using wireless sensonrs"
for residential areas that rely on water storage tanks. A pH sensor and a total dissolved solids meter are
used in this system to provide accurate readings of the water's pH and dissolved solids levels. K-Means
clustering, a machine learning technique, has been used to forecast water quality using a training data set
66
ISBN: 979-8-3503-4821-7
consisting of various water samples. This system is now in its prototype stage, used with low-cost
embedded devices like the Arduino Uno and Raspberry Pi3.
Keywords-- IoT, Monitoring system, K-Means, Arduino Uno
PAPER ID : CS551
Functional overview of integration of AIML with 5G and beyond the network
Kolli Srinivas , Dr.Aswini.J , Dr. Pramoda Patro, Dr. Devendra Kumar
[email protected]
ABSTRACT : 4 G/LTE mobile networks resolved this issue. Strong physical layer and adaptable network
design enable high-capacity mobile broadband Internet. Despite this, the prevalence of bandwidth-intensive
technologies like virtual reality, augmented reality, and others has grown. In addition, the rising popularity
of new services places a strain on mobile infrastructure. Applications requiring high availability and low
latency, such Internet-of-Vehicles or communications between vehicles (IoV). With the advent of the new
5G technology with its massive MIMO radio interface, these problems are no longer a concern. Networks
protected by software-defined networking (SDN) and NFV have added a new level of flexibility that allows
network operators to serve services with very high requirements across several industries. Network
operators must increase and diversify their intelligence to fully comprehend the operational environment,
user behaviours, and user demands. A further goal is to become (self-) networkable proactively and
effectively. This chapter will look at how AI may help us in the modern world. Next-generation mobile
networks that are both efficient and adaptable may benefit greatly from machine learning in the 5G era and
beyond. The evolution of AI and ML in network applications.
Keywords: 5G, Artificial intelligence, Machine learning, Neatwork controlling system
PAPER ID : CS554
A Machine Learning Approach For Predicting Crop Seasonal Yield and Cost For Smart Agriculture
Nandaraj M, Roshan Raj A, Uma Maheshwari M, Sathiyaraj R, Tejasvi K
[email protected]
ABSTRACT : In our economy, agriculture plays a key component.
Current scenario has plunged the agricultural sector into extensive losses leading to a lot of food shortages
and cases of
farmer suicides. This can be solved through the implementation of advanced scientific methods like
Machine learning, Deep Learning, and Internet of Things. The Proposed framework Smart Agriculture
using Seasonal Yield and Cost Prediction (SYCP) uses Machine Learning to predict the ideal crop or set
of crops to be grown based on a particular season. The Model also analyzes the current market trends and
predicts the approximate price of the crop for that season and geographic region based on which the farmer
can decide accordingly. Random forest has been found to be suitable for both crop and price prediction.
The framework has been experimented with an agricultural dataset and the results were found to be more
efficient than the existing methods. This would be an efficient solution, where it offers a seasonal yield
along with price prediction to improve the economy of farmers and to push them to grow.
Keywords: Agriculture, Machine learning, Seasonal yield, Cost prediction, Random Forest.
67
ISBN: 979-8-3503-4821-7
PAPER ID : CS557
One-dimensional convolutional neural network and long short-term memory convolutional neural
network for 12-lead ECG arrhythmia detection.
S.V.Evangelin Sonia, Dr.R.Nedunchezhian, Dr.M.Rajalakhmi
[email protected]
ABSTRACT-- Many researchers have spent years trying to create an effective algorithm for identifying
cardiac arrhythmias. However, despite the fact that they have yielded satisfactory outcomes, they have only
been used on either relatively small or uniform data sets. In this work, we provide the results of a study
designed to detect cardiac arrhythmias in 12-lead ECG signals of varying lengths, as collected as part of
the PhysioNet/Computing in Cardiology Challenge. We came up with a method that begins with looking
at the database’s labels. Following this, a number of signal processing methods were used on the 12-lead
data to remove noise. Next, the data were segmented and filtered to determine which beats were most
crucial in demonstrating the arrhythmia. Then, a 1D-CNN+BiLSTM model was trained and evaluated on
the training set and not only on the training set but also on the hidden validation and testing sets, all using
a 10-fold cross-validation scheme. In the validation set, our method scored 0.854 as accuracy, and in the
entire test set, it scored 0.812.
Keywords-- Neural Networks, CNN, LSTM, ECG, Cardio vascular disease
PAPER: CS558
E-Commerce Customer Churn Prediction Scheme Based on Customer Behaviour Using Machine
Learning
Nagaraj P, Muneeswaran V, Dharanidharan A, Aakash M, Balananthanan K, Rajkumar C
[email protected]
Abstract— Customer retention is crucial for any company to succeed given the rapid growth in the number
of companies in established sectors. To analyze and research client retention, numerous approaches (such
as data mining and statistics) have been offered. Building digital (CRM) which is known as Customer
Relationship Management systems is a new trend that is growing in the global economy as a result of the
explosive rise of digital systems and related information technologies. In the telecoms sector, where
businesses are progressively going digital, this trend is especially pronounced. Predicting customer attrition
is a key function of contemporary telecom CRM systems. Keeping consumers has grown increasingly
crucial for insurance businesses, and causes for churning are challenging, according to competition in Iran's
insurance industry in recent years and the entry of the private sector. Churners have continually been a
serious drawback for each business that provides services. Churning drives up a company's expenses
whereas additionally lowering its gross margin. Customers WHO request service termination are usually
exhibiting client attrition. those that offer the information keep in government databases, furthermore
because the organizations WHO fund the gathering of such information, are at the same time turning into
a lot of conscious of however tools that enhance analytical capabilities additionally offer threats to the
privacy of information records. However, it is the potential to forecast in case a customer wishes to avoid
the service using predictive analysis according to previous utilization, the performance of the service, and
other kinds of patterns. Customer churn is a problem for many sectors, but it is especially severe in the
fiercely competitive and currently widely liberalized mobile telecommunications sector. The average
monthly churn count for mobile telecommunications is reportedly 2.2%. In addition to potential costs from
lower revenues, losing customers also increases the demand for acquiring new ones.
Keywords— Customer retention, Churn prediction, ECommerce, retention rate, predictive analysis
68
ISBN: 979-8-3503-4821-7
PAPER ID :CS560
Smart Home Automation Security Using Type-2 Fuzzy-Based Approach through IoT
Dr.T.N.Aruna
[email protected]
ABSTRACT: Background and Objective:
Take care of data with security condition is a key predicament in residential and undersized areas. The
majority of this area allocates a substantial piece of their finances to ensure their data are strong. Internet
of things method has been generally useful to detecting the sign of hack such as hack information and data.
This revise intend to expand a method that establish security condition Type-2 Fuzzy judgment and IOT
assess the functionality of the future indicative security condition. This revise add to the residential method
can help house in precisely make a safe while IOT indication has been confer in this chapter to finish full
security for home using Type-2 Fuzzy. Methods: This loom augment the routine of our effort as it assist
insecurity probability gives result exactly. The security information is get starting Type-2 Fuzzy convention
are applied to calculate the odds of home in as short, average or serious problem. By concern Type-2 Fuzzy
method to IOT information which necessitate to be procedure choice creation in home security. Results:
Internet of things and Type-2 Fuzzy method and purpose are worn to recognize the stage of danger issue
to get safety measures in proceed to keep their home safe. Conclusions: Complete reasonable revise gives
an appropriate platform for support revise by ruling the work of Type-2 Fuzzy logic using home safe.
Keywords: home security, ambiguity, Type-2 Fuzzy Reason, Data security, Internet of things.
PAPER ID :CS561
Classifying the status of CKD using Randomized Weighted OptimizationModel Using Machine
Learning Techniques
ABSTRACT— chronic kidney disease (CKD) is a substantial health-care burden owing to its
growingoccurrence, enhanced danger of development to end-stage renal disease, and reduced mortality and
morbidity prognosis. It is quickly accelerating to a worldwide well-beingcatastrophe.The main causes of
CKD include high blood pressure and diabetes. The major function of the kidney is filtering waste from a
human body. When kidneys fail, waste builds up in bodies and eventually leads to death. Researchers all
around the globe utilize the Glomerular Filtration Rate (GFR) and kidney damage indicators to define CKD
as a disorder that causes decreasing renal function over a period of time.A person having CKD is more
likely to die early. Doctors have a tough time recognizing the several disorders associated with CKD early
enough to avoid the condition. Data mining approaches have recently been exposed to considerable
research in CKD diagnosis, with a focus on accuracy, either through the simplicity of illness by conducting
feature selection in addition to pre-processing or not before classification.This research work determines
creatinine level is a kind of blood metabolite that has a significant relationship with GFR and Blood Urea
Nitrogen (BUN)BUN Creatinine Ratio (BCR). Since measuring GFR and BCR is challenging, this work
focused in analysing creatinine value is used to determine Estimated GFR (EGFR) and BCR from serum
creatinine indirectly utilizing the attributes available in the dataset whereas the 28 attributes are considered
inclusive of “Gender” with 523 records. The EGFR and BCR computing results areused in this prediction
analysis for providing an accurately categorizingthe statusof CKDwhich has been utilized by Randomized
Weighted Optimization Model (RWOM).Furthermore, the proposed RWOM with Neural Network (NN)
is compared to the RWOM with Logistic Regression (LR), NN, LR in terms of analysing the better
categorization status of CKD from patient records.
Keywords— CKD, eGFR, RWOM, BUN, creatinine, non-CKD, classification
69
ISBN: 979-8-3503-4821-7
PAPER: CS562
FundusDBNet: Methodology to detect Retinal Fundus Diseases
Rishikesh Sivakumar, Nidhi Singh, Dr.C. Jothi Kumar, Dr.N. Prasath
[email protected]
Abstract— Multiple eye diseases can be easily detected from the orientation of the fundus and the blood
vessels on the fundus of the eye. Most of the already present methodologies use methods like Gaussian
filter, channel extraction and other preprocessing techniques for blood vessel segmentation. Which does
not really solve the problem of giving faster and better results. Our approach involves data augmentation
on the obtained samples as the number of samples is less and augmentation of these samples would result
in better performance and efficiency of our model. We use an Encoder-Decoder Model, which is coded for
segmenting the blood vessels from the samples. Hence providing more clear and accurate segmentation.
Deep Learning Models like VGG-16, VGG-19, and ResNet50 can be applied to the blended images for
predicting if the patient suffers from Retinal Disease or not. Since we use deep learning techniques and
complex models, highly accurate results are obtained. This could be a very efficient alternative to reduce
the efforts of an ophthalmologist and a major step towards automation in the field of ophthalmology.
Keywords- segmentation; disease detection; retina; deep learning; computer vision, U-Net.
PAPER ID : CS564
Cervical cell classification using Deep Learning Techniques
Remya R ,Kumudha Raimond
[email protected]
ABSTRACT— Cervical cancer is a very common disease in women. Early detection and accurate
diagnosis can save lives. Pap smear or pap test is the widely used technique for the early detection of
cervical cancer. Manual analysis is time consuming process that relies on expert cytologist. Hence it is
better to develop a computer aided system for the diagnosis. Depending on the level of abnormalities, the
cells in papsmear cluster slides vary in colour, size, and morphology. Various machine learning and deep
learning techniques are developed for the segmentation and classification of the pap smear image. In this
paper five different deep learning networks Vgg-16,vgg-19,Resnet-50 and ResNet-101 and Efficient Net
v3 are compared on the SIPAKMED dataset. The architectures were also evaluated on the Herlev
benchmark dataset.
Keywords- Deep Learning, cervical cancer, transfer learning
PAPER ID : CS567
Application of Latest Authentication Technologies in Vehicular Cloud
J.Varalakshmi, Dr.S.Dhanasekaran
[email protected]
ABSTRACT— Vehicular Network technology has emerged over last few years under Intelligent
Transportation system. Due to the growth of the massive amount of information exchanged in VANET,
Vehicular cloud is a new paradigm in which all data are stored in cloud servers. Security is an important
issue in vehicular cloud. Many researchers had proposed many schemes to resolve the issues related to
authentication, confidentiality, Non-repudiation, availability, reliability and other performance parameters.
Still we are facing many challenges in authentication mechanisms of Vehicular cloud. This paper will
investigate many different challenges in authentication schemes on VANET, cloud computing and current
trends. This paper focuses on study of Challenges on mutual authentication schemes in VANET, cloud
environment and Vehicular cloud. Finally, this paper will also discuss the current authentication
technologies so that it will make the researchers to progress towards the successful deployment of these
70
ISBN: 979-8-3503-4821-7
technologies in Vehicular cloud so that it will be useful to resolve any security issues and implement a
novel and an efficient authentication protocol in vehicular cloud in future.
Keywords- VANET; Vehicular cloud; Authentication Technologies; Blocking ID Mechanism
PAPER: CS568
Adaptive Energy Routing Protocol using Spider Optimization in Wireless Sensor Networks
Dr. A.Srinivasa Reddy, G.Malleswari
[email protected]
Abstract— A WSN is a collection of sensor nodes that monitor and record the condition of the physical
atmosphere and compile the obtained information in one place. The Next Generation Networks, which
include the Internet of Things, Sensor Control Networks, Ubiquitous Sensor Networks, and Machine-
Oriented Communications, use WSN technology as their fundamental foundation. To complete the sensing,
communication, and processing functions of WSN, an energy-efficient routing protocol is required to
manage the network energy dissipated and to reduce traffic and overhead during data transmission phases.
To balance the energy consumption of each Sensor Node in the network, clustering is utilised as a method.
The Firefly and Spider Optimization algorithms serve as the foundation for the ensemble bio-inspired
technique we offer here. The suggested protocol prevents unnecessary routing messages, which could lead
to significant energy waste, by frequently recycling information from the source node into the sink. The
ideal route path can be chosen by this routing protocol. The suggested algorithm made use of several
characteristics, including node residual energy, inter-cluster distances to the sink, and cluster overlaps, to
choose the optimal cluster heads at each round. To achieve the best performance for the network
requirements, the parameters of the projected solution can be adaptively modified during the clustering
process.
Keywords—Sensor Node, Firefly, Spider Optimization, Routing Protocol, IoT, Sensor Control Networks,
Ubiquitous sensor networks.
PAPER ID : CS569
A Survey on existing system design used for managing ambulance booking through mobile App
Dr. L. Srinivasan, Ms.Mounika BR, Ms.Shaik Anju Minayar, Ms. Sanjeevani Sharma , Student
Ms. Navya Sood
[email protected]
ABSTRACT— In today’s world and in India, due to rapid growth in population and increase in traffic
accidents are bound to often as the clock runs around.our goal is to design an mobile application to assist
the needed patience with medical care.The motto of the design is to bridge time line between the ambulance
arrival and the call made by the patient.. Due to the heavy traffic, the ambulance driver initially had no idea
of the precise location of the accident scene. As a result, we were not able to save many lives. Since
everything in the world today runs on smartphones and apps, we have developed a mobile application that
allows real-time monitoring of ambulance services. Ambulance drivers will record their availability and
location using this app.In this project, the Ambulance Mainframe System Android app has been defined.
The most popular rescue service, 1122, can be contacted by phone, but booking an ambulance with an
Android smartphone is a whole separate concept.The Boosted App forms an ambulance request, which is
immediately updated on the assembly mainframe office, where a 24/7 server will automatically check the
coordinates of the request and respond to the user and several nearby stations. The aforementioned
smartphone app could very well change the way people use ambulance services by starting to make them
more convenient and reliable for such a small department. His app for an injured person seems to work
with a single right click, which passes notifications/requests to emergency services, such as through a
71
ISBN: 979-8-3503-4821-7
general packet radio service between ambulance service vehicle owners, with user and location information
stored inside the file (administration system).
Keywords- Ambulance size, pick-of-point, destination And location, google maps, public safety.;
PAPER: CS570
A Review on Techniques used for Face Authentication based Smartdoor Bell System using IOT
Dr .L. Srinivasan, Naveen Kumar Reddy, Basavaraj U P, Darshan M E, Chirag K
Abstract— The Internet of Things (IOT) is the correspondence between different elements: devices,
designs, vehicles, and equipment, programming, sensors, and the accessibility of frameworks. An IOT-
aware theft order provides the system. The Internet of Things is based on high-level human-machine
correspondence that approximates machine-machine correspondence. The motivation behind this venture
is to make home or office or any region secure. At the point when somebody presses the doorbell, then, at
that point, the doorbell settle on a video decision to the enlisted number. Assuming that somebody meanders
before the entryway it advises you by sending message. Then he can see the individual who is wandering
before our entryway. In this way, if the individual is realized we can open the entryway if not we would be
able be ready. And furthermore we can converse with the individual through versatile just and the individual
can answer there itself, Since it contains the sound speaker with the goal that we can hear the outside
individuals talks box the portable once we get the video call. In the event that somebody attempts to take
it, the take caution will been acted.
Keywords: Smart Doorbell, IOT Doorbell, Security IOT, Smart doorbell, Smart lock, Wireless doorbell
PAPER: CS571
Effect of Supervised Machine Learning Algorithms on Rice Plant Diseases
R. Sahith, Dr. P.Vijaya Pal Reddy
[email protected]
Abstract— As one of the top five producers and consumers of rice in the world, India largely relies on rice
to support its economy and satisfy its nutritional demands. To guarantee the healthy and proper growth of
the rice plants, it is essential to identify any illness early on and before the damaged plants receive the
required treatment. Having an automated approach makes sense because manually detecting diseases takes
a lot of time and effort. This article describes a method that uses machine learning to identify illnesses in
rice leaves. Four of the most common diseases impacting rice plants are Blast, Bacterial Leaf Blight,
Tungro, and Brown Spot, according to this study. The input consisted of clear pictures of damaged rice
leaves. Following the appropriate pre-processing, a variety of machine learning algorithms, including
Random Forest, J48, REP Tree was trained on the dataset. After 10-fold cross-validation, the decision tree
method produced results on the test dataset with an accuracy of above 94%.
Keywords—Random Forest, J48, REPTree, Performance measures, Rice diseases.
PAPER ID : CS575
Farmer’s Friendly Robot using Internet of Things
Florance G, Aadhithiya R K, Santhosh Krishna B V, Anuraag Ramesh, Srajan Sharma
[email protected]
ABSTRACT— The agriculture sector is one that requires a lot of labor and resources. As a result, farmers
are increasingly relying on automation and technology to solve this problem. The cost, speed, and
complexity of agricultural robots prevent them from being made accessible to the public. Therefore, the
72
ISBN: 979-8-3503-4821-7
plantation sector has a slow adoption rate for new industrial advancements. The creation of a low-cost
agricultural robot for irrigating fields with water, insecticides, and seeds is described in this research report.
The plantation robot created for this research project is primarily used for three tasks: showering crops with
pesticides, water and also seed sowing. It is a four-wheeled robot with a wireless controller for controlling
it, a camera that broadcasts a live video feed for monitoring crop health.
It even functions as a robot land analyzer that can compile vital data about the soil, send it to the farmer
over an IoT network, and store the data. It even looks at the information to make future decisions that could
help in making crops healthier. It is an automated, multi-tasking robot that shall monitor the terrain and
behave appropriately as and when needed. The robotic vehicle periodically scans the land and keeps an eye
on the data after a set amount of time. Farmers can quickly determine if irrigation is necessary. Last but
not the least, the robot's remarkable battery life guarantees that there shall not be a rise in operating times
or a decline in the efficiency of the pesticide spraying process because human personnel need to be replaced.
Keywords- Internet of Things, Seed sowing, Pesticide spray, Crop health, Water Irrigation, Pumps,
Microprocessors
PAPER: CS577
An Improved System for Students Feedback Analysis Using Supervised Probability Techniques
Sreedhar Kumar S, Syed Thouheed Ahmed, Greeshma Sreejesh, Afifa Salsabil Fathima
[email protected]
Abstract— Sentiment analysis is the process of detecting positive or negative sentiment in text.
Educational institutes and universities are showing interest to improve quality of education system by
monitoring teacher’s, teaching, student’s learning, and course analysis using feedbacks. It aims to train the
existing unstructured feedback and to classify the sampled feedback belongs into positive, negative or
neutral polarity based on probability technique and supervised concepts. It consists of four stages input
comment set, pre-processing, feature extraction and classifier respectively. In the first stage, the raw
datasets are collected from server by N number of user feedback data X. In the stage two, the system
reduces number of words in each individual text command to recognize the significant and insignificant
words and eliminate the insignificant words based on predetermined model. In the third stage, It aims to
reduce the number of features in a data set by creating a new feature from the existing one. In the next
stage, the positive, negative and neutral keywords are forwarded by feature extraction and keyword input
dataset are mixed with all possible key value based on Naïve Bayes approach and training data. The
Experimental result shows that the proposed approach is well suitable to train the unstructured feedback
and classify them into positive, negative and neutral polarity with better clarity.
Keywords—Sentimental Analysis, Students Feedback, Naïve Bayes, Probability based text classifier,
Classifier, Predetermined model
PAPER ID : CS579
HAND SIGN CLASSIFICATION TECHNIQUES USING SUPERVISED AND UNSUPERVISED
LEARNING ALGORITHMS
J L V Sri Harshitha, Srithar S, N Saranya, J Vijaya Chandra, Yekkitilli Sruthi
[email protected]
ABSTRACT : Gesture recognition is becoming a sore subject in computer vision captivating the idea of
social interaction not only between humans but also create ergonomic systems that control devices ranging
from time-of-flight cameras and controlling vehicles to virtual reality. Even though it is gaining ground
recently, fast and robust recognition remains an unsolved problem. To our understanding, we are utilizing
73
ISBN: 979-8-3503-4821-7
the Leap motion sensor captured near-infrared image dataset for gesture identification, which tides over
word-level sign recognition encompassing a diverse set of hand gestures. We have suggested the approach
of extracting the gesture from the image using PCA and image segmentation followed by the feature
extraction stage. In this paper, we have analyzed and differentiated various methods of hand recognition
counting Random forests, stochastic gradient descent, Naive Bayes, Decision tree, and Logistic regression
algorithms and exploited our results.
Keywords— Gesture, Recognition, HCI, Segmentation, Machine learning, Accuracy.
PAPER ID : CS583
Relating the DT Framework in Securing the Healthcare Data using Blockchain
Dr.M.Shobana, A.Mukeshkumar, Dr.S.Vaishnavi, Mr.C.Gokul Prasad
ABSTRACT - In this digital world, everyone is required to visit a hospital to resolve illnesses or medical
conditions. there are 87% to 95% of clinics or multi-specialty hospitals in the world use a hospital
management system to maintain the hospital for finance and analytics of patients and doctor details. there
are problems with storing data or details about doctors or patients or hospital information in a centralized
server. this server may be hacked, or a virus attack may occur through data leakage or web support. if any
software is installed by any person or server officials from an unknown or malicious website to increase
usability, it may cause a virus attack or hacking. The essential documents for this event have been leaked
or corrupted, which may cause critical situations for patients and doctors. to avoid such problems, we
present the solution as software with blockchain technology called "BHealthy" which has been created with
all stages of Design Thinking. It will be secure and stored in encryption format because it runs on a
decentralized server and blockchain technology. It is unaffected by virus attacks or hacking. by using this
method, data is safe, and doctors and patients do not need to be concerned about details storing format.
Keywords: Healthcare, Decentralized, Interplanetary file storage, Blockchain Network, P2P Network,
Design Thinking (DT)
PAPER ID : CS585
V-VERIFY WITH BIOMETRIC SYSTEM USING DESIGN THINKING FRAMEWORK
Dharshana R, Monica V, Keerthana S, Swathi G, Pranava Kailash S P
ABSTRACT— This The most common way of voting in our nation is ELECTRONIC VOTING
MACHINE. It was introduced to ease the process of voting system in our country. But the difficulties the
electoral commission faced in the verification of the voter remained the same. The verification system that
is prevalent in the nation is paperwork. The officers from the electoral commission have to verify the voters
and their voter ID card with the documents they have and it is time consuming for both the voters and the
officers. To have a fast and easy verification system, this paper proposes a novel approach. The EPIC
number of the voter will be entered and the database will be cross verified. The next step is to verify the
fingerprint and the face of the voter. If the fingerprint and the face match the voter can cast their vote. The
facial recognition is achieved through the OpenCV and the Finger print recognition uses Binarizing and
Ridge Thinning to identify the genuine fingerprint from the database.
Keywords- Voting, Verification process, Fingerprint recognition, Facial recognition, Arduino UNO,
Design Thinking
74
ISBN: 979-8-3503-4821-7
PAPER: CS588
AI - UNDERWATER DRONE IN PROTECTION OF WATERWAYS BY RELATING DESIGN
THINKING FRAMEWORK
SHOBANA.M, MADHAVAN.M, MEHA V.S , NANDHINI.S, NEERAJ.D
Abstract : A new type of drone is being created, with the aim that it will assist in the exploration and
protection of our waterways. This AI-powered drone which is selected through a new problem solving
technique Design Thinking would be able to navigate and explore the undersea environment without the
assistance of a person. The drone will be equipped with cameras and sensors to collect data and information
under the water environment. Water is a basic necessity for all living creatures; it is necessary to keep water
hygienic and clean. Water becomes polluted for a variety of reasons, including industrial waste, rubbish
waste, and so on. Conventional means are used to purify water from lakes and ponds. Drone movement
will be controlled by artificial intelligence, and targets will be identified through object detection. It
includes a night-vision camera as well as lighting. The drone is attached to a buoy that warns seafarers and
is equipped with a solar panel and garbage collector, allowing it to clean contaminated waste that is deep
under water.
Keywords : Underwater, Artificial Intelligence, Drone, Detection, Clean water.
PAPER ID : CS589
NEURAL NETWORK-BASED SOFTWARE VULNERABILITY CLASSIFICATION TO
ENHANCE ACCURACY
Dr. M. Jagadeesan ,Dr.T.M Saravanan, Dr.P.A. Selvaraj, V. Mounika ,D. Ramya ,K.S. Vasantha kumar
[email protected]
ABSTRACT— The security risks are increasing due to software flaws. The vulnerability dataset used in
this design had taken from the cve web site. Additionally, the updated error description is matched as input,
cross-referenced with the dataset and if a match is found, the error order is set up. It is now possible to
manage software vulnerability using an automatic bracket system, and the system's security performance
will also improve. It will also reduce the chance of system damage and attack. In this design, a brand-new
model called the automatic vulnerability bracket model built on bracket algorithms for Convolutional
Neural Networks has been put out. 900 neuron input layers, 16 retired layers, and 27 affair neurons are
used to build the model. Age range is adjusted, and it is delicate.
Keywords: software vulnerability, Machine Learning, SVM, k-nearest neighbor, Deep Learning,
Convolutional Neural Network
PAPER ID : CS591
Prediction of Mango Leaf Diseases using Convolutional Neural Network
Dr T M Saravanan, Dr. M. Jagadeesan, Dr. P. A. Selvaraj, M. Aravind, G. Dharun Raj, P. Lokesh
ABSTRACT— The Convolutional Neural Network CNN works by obtaining a picture and designating it
with some weightage supported by the various objects of the image, to distinguish them from one another.
CNN needs little or no pre-processing information as compared to different deep learning algorithms. Early
diagnosis and correct identification of mango plant disease prediction will manage the unfolding of the
diseases Mango leaf diseases damage mango quality and yield.This research uses deep learning to
automatically identify leaf diseases in different mango plant kinds. The planned work is Associated with
the Nursing correct identification approach for the mango plant disease prediction exploitation of the
Convolutional Neural Network. It includes generating comfortable method pathological pictures Associate
in nursing coming up with a model and a design of the Convolutional Neural Network to discover mango
leaf diseases. The image augmentation method is employed to extend the number of images. completely
75
ISBN: 979-8-3503-4821-7
different information augmentation techniques square measure applied to stop overfitting and improve
generalization.
Keywords- Mango leaf, Image preprocessing, Convolutional Neural Network.
PAPER ID : CS593
Recognition and Processing of phishing Emails Using NLP: A Survey
Chunduru Anilkumar, Nandam Sri Mouli , Aravind Karrothu, Chaduvula Bhanu Tej
[email protected]
ABSTRACT— Email is preferred by both individuals and businesses as one of the most important forms
of communication. Despite the popularity of alternatives like electronic communications, mobile
applications, and social networks, email usage and importance have been raising continuously. Phishing is
the fraudulent practice of stealing any person’s confidential information, such as financial details, login
information etc. In which they look completely authentic and genuine and extremely hard for victim to
detect whether the mail is spam or not. Moreover, the number of phishing emails is growing daily. So,
handling of phishing emails is a crucial research issue. NLP techniques are useful for training data to detect
spam content. In this work we studied leading research areas in phishing email detection and features used
in spam emails. Feature extraction and selection plays a key role in phishing email detection. This paper
also explains how actually the phishing attack happens and obtained results including information about
concepts of natural language processing and working procedure.
Keywords— Phishing Emails, Natural Language Processing (NLP), Machine learning, Email classification,
Spam Detection
PAPER ID : CS599
ADVANCMENT IN DRIVER ASSISTANCE SYSTEM USING IOT TO PREVENT ACCIDENTS
Dr.A.Sumithra, Shakthi Shree.S.K, Sherlin Mary Beaulah.S, Sindhuja. M, Smirna Martin
[email protected]
ABSTRACT— In this paper, we are to discuss the High level Driver Assistance System. High level Driver
Assistance Systems have been quite possibly of the most dynamic region of ITS examinations over the
most recent twenty years. ADAS plan to help drivers by all things considered giving admonition to diminish
risk openings, or mechanizing a portion of the control undertakings to free a driver from manual control
from a vehicle. ADAS capabilities can be accomplished through an independent methodology with all
instrumentation and insight on board the vehicle, or through a helpful methodology, where help is given
from streets and additionally from different vehicles. In this article, ongoing examination and
improvements of longitudinal control help frameworks are assessed including versatile voyage control,
forward impact cautioning and evasion, and platooning collaborators. The audit centers around contrasting
between independent frameworks and agreeable frameworks concerning innovations utilized, framework
effects and execution. The fundamental goal is to accomplish normal figuring out on ADAS utilitarian
possibilities and limits and to distinguish research needs for additional examinations.
Keywords- Automation, Switching, Energy protection, ArduinoSensors
76
ISBN: 979-8-3503-4821-7
PAPER ID : CS601
COMPARATIVE PERFORMANCE ASSESSMENT OF MACHINE LEARNING ALGORITHMS
TO PREDICT CARDIOVASCULAR DISEASE
S.Hemalatha ,Dr.T.Kavitha, D.Niruba, S.Nandhakumar , R.Venkatesh
[email protected]
ABSTRACT-- Cardio Vascular Disease (CVD) are a major factor in many deaths across the world. Many
people in their middle or later years suffer from heart conditions, which typically result in significant health
problems include heart attacks and strokes. In order to stop any significant health problems before they
arise, it is vital to detect and anticipate cardiac illnesses. In order to predict cardiovascular diseases at
various evaluation stages, this paper used a variety of machine learning approaches, including Neural
Networks (NN), Decision Trees (DT), Naive Bayes (NB), Logistic Regression (LR), Random Forest (RF),
Support Vector Machines (SVM), K-Nearest Neighbor (KNN), and XG Boost. Results reveal that the
Random Forest approach performed better than the other methods and had a predicted accuracy rating of
98.25%.
Keywords ML, CVD, DT, RF, NN, NB, LR, SVM, KNN.
PAPER ID : CS603
A Visionary Approach to Detect Spoofing Website using Machine Learning Algorithms
Dr.T.Kavitha , S.Hemalatha, R.Mounica3, V.Niveda4 , Yasvant Kumar5
[email protected]
ABSTRACT - Today, everyone is addicted to the internet. Web searching has become important for
everyone despite their age, location or occupation. However, as the use of the internet is increasing, the
spoofing attacks through malicious websites are also increasing. All conducted online purchases and online
pursuits being online amount transaction,e-reservation, recharge online and more online. As the number of
Internet users grows, automatic URL detection is becoming increasingly more important for protected
users. The proposed research will compare the contrasting classification models such as Random Forest,
Decision Tree and Logistic Regression which have the highest accuracy of differentiation between
legitimate sites from phishing sites. The intention of the research is to find the attacking sites and best ML
algorithm which has the highest accuracy.
Keywords: URL; machine learning,Logistic Regression, Decision Tree,,Random Forest
PAPER ID : CS605
ARTIFICIAL INTELLIGENCE-BASED HEALTH CARE MONITORING AND RECORD
STORAGE SYSTEM
Sathiya Lakshmi E , Gomathi R M ,Meenakshi K, Brumancia E , Nithya M
[email protected]
ABSTRACT - The expanding integration of Machine Learning (ML) and Artificial Intelligence (AI)
towards medical care raises the risk of direct harm through modeling bias; still, partiality across many
biomedical AI applications remains unmeasured. This paper proposes a structure to objectively evaluate
medical AI from a variety of perspectives, with an emphasis on binary classification methods. The models
were assessed prospectively and sequentially using conceptual measures of discrimination, accuracy and
reliability measure of individual-level error. In the prediction models, we found multiple instances of bias
at the model level. Individually, we found that most models had slightly greater margins of error among
older people. A model can be biased towards some identifiable features in specific tasks while being biased
towards another protected group. Therefore, current biased assessments may not provide a full
representation of the varying effects of a model on its subgroups. Only a complete examination with a
77
ISBN: 979-8-3503-4821-7
rigorous search for undiscovered bias may yield adequate data for an unbiased assessment of AI that needs
to follow by discovering the fundamental causes of bias, and finally, affecting changes.
Keywords- E-Health records; Healthcare facility; Performance Measures; Pandemic disease
PAPER ID : CS606
Prediction of Brain Tumor Classification by using CNN
Dr. Anitha Julian, A V Charan Reddy, B Naga Narendra, Chiranjeevi A
[email protected]
ABSTRACT- An abnormal cellular mass that forms in the brain is called a tumor. Brain tumors come in
a wide variety of forms. Different types of brain tumors may be classified as either benign (non-cancerous)
or malignant (malignant). Brain tumors may originate in the brain itself or can be the result of metastasis
from cancer in another section of the body (secondary, or metastatic, brain tumors). Treatment for brain
tumors may range from observation and observational care through surgery and radiation therapy,
depending on the tumor's characteristics and location. Biopsy is used to classify brain tumors; however, it
is not typically performed prior to decisive brain surgery. Radiologists can diagnose tumors without using
invasive procedures thanks to advances in technology and machine learning. Convolutional neural network
is a machine-learning technique that has produced notable results in picture classification (CNN). If more
feature extraction techniques are added to assist the CNN method and correctly categorize images of brain
tumors as normal and abnormal, it is projected that the success of the results would grow.
Keywords- Deep Learning, TensorFlow, Keras, CNN.
PAPER ID : CS608
Automatic video summarization and classification by CNN model: Deep learning Surendra Reddy
vinta, Dr. Pavitar Parkash Singh , Dr. Ajoy Batta ,N. Shilpa
[email protected]
ABSTRACT— As smartphones and other camera-enabled devices become more mainstream and user-
friendly, more people are recording and sharing films through social media and video streaming websites.
This makes them an essential tool for spreading information. It's a pain to watch and evaluate so many
movies. An automated video summarizing gives a concise analysis of the source material, which is useful
for indexing and categorizing long films in the video database. Putting together a synopsis for a video is an
uphill task. By simulating a two-stream architecture with a deep convolutional neural network in each
stream to extract a video's spatial and temporal components, this research hopes to automate the process of
making video summaries. Video segment highlight scores may be generated using a two-dimensional
Convolutional Neural Network (CNN) that uses spatial information.
Additionally, a 3-D convolutional neural network (CNN) includes temporal data. The ratings for each
segment in each stream are averaged to determine which portions of the video are the most compelling.
Since the highlight result only conveys a relative degree of interest, the DCNN in each stream is trained
using a pairwise deep-ranking model. With some model tweaking, we can make the highlighted part of the
video score higher than the rest. Videos summaries may be created from the retrieved clips.
Keywords- Deep learning, 2-D DCNN, Video summary
78
ISBN: 979-8-3503-4821-7
PAPER: CS619
BRAINY BIN FOR METROPOLITAN CITIES USING DESIGN THINKING METHODOLOGY
Dr B.Vinodhini, T.V.Madura, M.Mohammed Sufaid, Pavatharani Shanmugamani, B.L.Pavithra
[email protected]
Abstract - With the tremendous rise in population, a huge number of wastes is getting piled up at different
places in both rural and urban areas. Overflowing bins in public places leads to an unhealthy environment.
This unhygienic society creates a way for the easy spread of diseases to the people in that locality. Thus,
Cleanliness becomes the need of the hour. Proper Waste Management has to be created for obtaining dirt
and germ-free society, Where the people nowadays often end-up violating the Waste Management Rule
which is proposed by the Government Of India in 2016. This problem caused due to two reasons when the
bins are filled and the people who wants to throw the wastes has put into the situation of throwing wastes
outside of the bins and another reason is misusing and throwing the wastes where-ever even though the bin
is empty. To overcome these issues which are increasing day by day, we need to apply some technology-
based. The recent technologies provide radical solution and effective way to handle the problems which
has been faced. In this project Wireless Sensor Network (WSN) and Internet of Things (IOT) technologies
(which includes the integration of several devices like ESP32 chip, GPS module with patch antenna, object
detection sensors, ESP32- CAM, Arduino board and RFID – Radiofrequency identification) are used for
efficient management of the waste bins and helps collecting fines from the people who violates the SWM
rule. This paper presents real time monitoring of the waste bins like bin level, location, live steaming of the
person who comes nearby for disposing the wastes, alert notification, SWM rule violation notification
through the result displaying in the Thing Speak application, It offers the capabilities of real-time data
collection, visualizing the collected data in the form of charts. and the alert notification can be viewed
through the LCD display as well. For reducing the man power and efficient management of the corporation
workers in evacuation of the filled bins timely at regular intervals the RFID reader is attached with the
embedded system and the RFID swipe card for the workers will be issued and they have to swipe and
authentication every time when-ever they came for evacuation purpose.
Index Terms—Cognitive Overflowing,unhealthy environment, efficient management, Prevention
PAPER ID : CS621
Improving Efficiency of Garbage Collection Vehicle for Smart Cities
Aviral Budhani, Mrs. Nidhi Sengar, Dr. Amita Goel, Mrs. Vasudha Bahl [email protected]
ABSTRACT: As the population of our country skyrockets, the consumption of goods has also increased
exponentially, we tend to face a chunk of problems thanks to all that. A problem that stands out of this is
the generation of enormous waste as the byproduct of this increased consumption. To tackle this inevitable
problem of waste management, the government has taken several steps in the right direction under the
Swachh Bharat Abhiyan scheme to create a clean, hygienic, and healthy country.One of the services that
have been started is the daily collection of garbage from every house. The availability of a garbage
collection vehicle was aimed at solving the problem of segregation of wet and dry garbage, as well as
biodegradable and non-biodegradable waste, and also help make the surroundings clean and hygienic.
However, the benefits came with a lot of shortcomings. The timing of the collection vehicle remains
uncertain. Causing a situation of hurry and sometimes leading to people missing the vehicle. Also, the way
people get notified that the collection vehicle has arrived is by the loud music that is played. The audio is
so loud that it becomes irritating so often. This is a cause of concern for places near schools, and hospitals
and also for doing work from home.To overcome these challenges, we tend to make a Mobile Application
addressing the above-mentioned issues. In this system, we are providing users as well as the drivers with
an app that will help them provide advance notification.
Keywords: waste collection, advance notification, Mobile Application
79
ISBN: 979-8-3503-4821-7
PAPER ID : CS623
SOCIALIZING PLATFORM FOR EDUCATIONAL INSTITUTIONS
Anitha Julian, Kishore R, Sathishkumar K, Alen Wenish J
[email protected]
ABSTRACT- The younger generation is increasingly setting the tone for popular culture, with an emphasis
on social network connection. Social media has evolved into a strong one-stop shop for all of our everyday
requirements. A social media website is a community where individuals may engage with one another
through chats or online calls, produce and share textual or video material, buy things, comment on other
users' postings, construct their own network of contracts, and much more. Everything is done online,
including communication, selling, and advertising. In this regard, creating a social media platform for
students to engage with the university is necessary for student career development. Social media provides
a smoother, more direct communication tool between students, teachers and the Institution. With the use
of social media network in Institutional level, a student can get to know about various events, club activities,
seminars, etc. which are taking place or about to take place in the institution. Additionally, the platform
improves students' academic achievement. Our solution provides a place for students to share photographs
of their numerous academic activities. This portal may also be used as a portfolio for all students by an
educational institution. Students can also use this platform to get to know about various events occurring
inside the institution. Collectively, this platform is a reliable and a feature-rich one. Keywords- social
media, socializing platform, communication tool, educational institution
PAPER ID : CS624
Soft Attention based Densenet for Severity Grading of Diabetic Macular Edema Mrs.V.Gayathri,
Dr.C.Bala Subramanian
[email protected]
ABSTRACT— Diabetic Macular Edema is a diabetic complication that leads to the blindness in working
age-population. DME causes severe damage to the blood vessels in the retinal tissue and leads to the
leakage of fluid and distorted vision. Diabetic retinopathy (DR) is considered to be a primary cause of
DME. The presence of DME is identified by the hard exudates (bright lesions) in the retinal images. We
can prevent the blindness in diabetic patient by the early detection of diabetic macular edema. There are
various methodologies for diagnosing and predicting the severity of the DME. The proposed work uses
Soft Attention based Densenet (SA -Densenet) for predicting the severity levels (focal DME & Diffuse
DME) of the diabetic Macular Edema. The proposed framework uses the kaggle fundus dataset for
predicting the severity level of DME. Throughout the experiment, measures including accuracy, precision,
and recall are assessed to check the performance of the proposed work. Comparatively the proposed
strategy is the most accurate in predicting diabetic macular edema from fundus images.
Keywords— Diabetic Macular Edema, Diabetic Retinopathy, SA-Densenet
PAPER ID : CS625
Analyze the political preference of a common man by using data mining and machine learning
Dr.G.Gokulkumari, Dr. Pooja Nagpal , Dr. M. Ravichand , Dr Renu Vij [email protected]
ABSTRACT- Using learning algorithms and game theory, provide a procedure for estimating the likely
political leanings of a given voting population. A combination of online and in-person polling was the first
step in our strategy. With voters' preferences for political candidates expressed via a linear hierarchy of
qualities, a zero-sum game strategy may divide the total number of votes fairly. In a case study of the
Major's election in our city, which took place in September 2020, we put our approach to work for electoral
80
ISBN: 979-8-3503-4821-7
prospection. The findings were quite promising. I suggest using this method in marketing campaigns for
genuine products.
Key words- political preference, data mining, voting
PAPER ID : CS628
EFFICIENT GESTURE-BASED PRESENTATION CONTROLLER USING TRANSFER
LEARNING ALGORITHM
Shaik Faiz Ahamed, Srithar S, Pala Sandeep, Paraselli Tushar
[email protected]
ABSTRACT-- This research work explains how to control Presentations using hand gestures.
Presentations are crucial in many aspects of our human lives. Everyone employee who has ever worked in
a company or been a student has participated in a demonstration at any point in their lives. PowerPoint
presentations might become irritating when you have to update and manipulate the slides using the
keyboard or a dedicated device. So, this project focuses mainly on the development of interactive projection
technology, which makes it more interesting and interactive by detecting the user's motion simultaneously.
Our team always wanted that users should be able to operate the presentation slideshow with hand gestures.
Many functions of the PowerPoint slideshow have been attempted to be managed in research using
gestures. This study attempts to map minute differences in movements using some simple PowerPoint slide
show control Python applications. The motions were recognized using machine learning. Gesture
recognition became more vital for operating software such as robotics, and games a few years ago. The
hand gesture recognition system improves when surroundings are in well-lit conditions. Hand gestures
make communication simple, useful, and technologically independent. The proposed technique is to help
presenters provide more effective and interactive presentations by increasing their natural contact with the
computer. The recommended approach, in particular, is more practical than a laser pen since the hand is
clearer and may attract the audience's attention more effectively.
Keywords— Presentation Control, Gesture, Transfer Learning, Classifier, Cursor.
PAPER ID : CS633
Safe Trade – A Stock Recommender using Machine Learning Algorithms
Anitha Julian, D V V Somasekhar , E GaneshReddy, V R Dikshit Reddy
ABSTRACT - The business and financial sector currently controls the majority of the global economy,
and stock market trading is a key activity there. The project uses Google's Tensorflow's Keras API to create
LSTM, training the algorithm on historical data to forecast stock prices. We focused on forecasting the
trend in a stock's value over the subsequent 20 days after the prediction day. It is a method where one
attempts to foresee the future worth of the company's present stocks to prevent loss or maybe make a profit.
With the use of several quantities that will be discussed later in this report, this study will show how
machine learning can be used to anticipate the same. To create somewhat accurate forecasts, we suggest a
machine learning algorithm that will be taught using various datasets of certain firms that were available
in the past. Shares market forecasting is a method for estimating the future value of a company's stock or
another financial asset traded on an exchange. The only goal of stock market forecasting is a measurable
substantial gain, along with, of course, avoiding large losses. Given that the efficient market hypothesis
states that these forecasts cannot be produced with the data that is now available, some people may question
the veracity of the findings of these predictions and conclude that they are fundamentally unexpected.
81
ISBN: 979-8-3503-4821-7
However, a variety of instruments and technology make it possible to learn about future trends, which
generates real income. The suggested approach detects frames with 85.5% accuracy.
Keywords: LSTM, ANN, CNN, Stock Market
PAPER ID : CS634
An Exploration of Evaluating the Performance of Malware Detection in the Cloud Environment
Charan Nivas Reddy Kaipu, Koneru Kamal Koushik, Gopu Karthik, Pavan Kumar T
Konala Sree Umadevi, Dr. S. Kavitha
[email protected]
ABSTRACT: Numerous businesses now adopt cloud computing, putting their server, storage, and
application resources on a network in the cloud somewhere online. Data from the user is stored on cloud at
a low price, making cloud computing the most rapidly developing technologies. In a nutshell, cloud
technology provides users with the access to massive infrastructures for low latency via a distinct
middleware that is comparable to Grid and HPC computing, both of which are currently in use. Over the
years, these kinds of systems have become more and more popular since they offer a setting in which to
host scalable applications. But rather than offering customers a common platform, the cloud network is
vulnerable to several security issues, including malware infections, DDoS assaults, SQL injection, and even
more complex attacks like zombie botnets. We describe our research on security challenges and virus
spread in cloud networks in this paper. However, the detection capabilities of conventional host-based
antivirus software are constrained, and these programs frequently miss threats; recent assaults are becoming
more complicated and evasive of conventional security measures. To understand cloud security and use
cutting-edge malware detection techniques to safeguard cloud, malware detection and research on these
approaches are described in this article. Because the resources in a cloud are delivered to the user in the
form of VM’s, they are susceptible to distributed denial of service attacks, malware exploits, and VM
Escape-based attacks. This paper discusses the various detection techniques and the most effective ones.
The discussion of each method will be covered from a broad perspective that also focuses on reliable
solutions, future directions, etc. The details are outlined in our article
Keywords: -- Cloud, malware, cloud security, databases, virtualization, virtual machines (VMs), cyber-
physical systems, malware detection
PAPER ID : CS635
An Image Based Classification and Prediction of Diseases on Cotton Leaves Using Deep Learning
Techniques
T Kalpana , R Thamilselvan, Dr T M Saravanan, M Pyingkodi, S Nandhakumar, V Priyanka, M
Sowndharya
[email protected]
ABSTRACT- Cotton is one of the most important crops but its cultivation is hampered by diseases and
infections. Rapid disease identification is the challenge due to lack of infrastructure. This necessities the
development of automated methods to predict diseases. So in this research Deep Learning Models (DLM)
could be used for the development of intelligent agriculture to effectively determine the spot of the diseased
cotton leaf part in the terrain. Semantic image segmentation is considered a major problem in computer
vision. So we investigate that deep learning of Convolution Neural Networks (CNN) for the deduction of
cotton leaf disease and implement the knowledge of deep learning models and transferring the knowledge
to our dataset. In this attempt we employed Image recognition model such as VGG16, Inception-V3, Image
data generator, flattening, batch normalization, conv2d and to extract diseased parts in cotton leaf using the
dataset of both healthy and diseased leaves. Optimization is used to choose optimal values and image
82
ISBN: 979-8-3503-4821-7
augmentation is done to improve the models. The result demonstrates that all trained models have an
accuracy more than 90% in classify cotton leaf disease. In particular VGG16, Inception-V3 attained an
accuracy of more than 96%.
Keywords- Deep Learning; Image Recognition Model; Inception-V3; Convolution Neural Networks;
Cotton Disease Recognition; VGG16.
PAPER ID : CS638
Hyperautomation-Beyond RPA:
Dr. R. Kavitha
[email protected]
ABSTRACT— In the technological ambience, the manufacturing industry sector must migrate to robotic
process automation, intelligent process automation, and artificial intelligence. By exploiting modern digital
corporate processes and technology breakthroughs, businesses all over the world are embracing automation
solutions capable of imitating human behaviors, eliminating repetitive work, and ultimately converting
employee roles into top-tier employment. After the pandemic, every industry education, manufacturing,
healthcare, etc., moved toward automation and internet activities. Most firms now rely on automation
techniques like robotic process Automation (RPA), Artificial Intelligence (AI), more recently Intelligent
Process Automation (IPA) and Hyperautomation to reduce costs, speed up processes, and improve the
accuracy of crucial business activities. The industrial sector has grown significantly and advanced
technologically over the years. Advanced technologies have always produced more efficient techniques for
converting raw materials into finished items. As a result, modern businesses have kept coming up with
novel uses for these tools to help automate manual processes and increase productivity across the entire
production line. Process automation in manufacturing refers to the application of technology to repetitive,
manual tasks that were previously successfully completed by people. Contrary to automation in other
industries, manufacturers are able to create products as they go through the production process using a wide
variety of machines and robots. Automation in this context typically refers to both doing the activity itself
and doing away with the need for human input during data entry. In the manufacturing industry,
"automation" refers to the use of machines to speed up systems or production processes. The ultimate goal
is to boost productivity by lowering costs or increasing production capacity, frequently both. Automation
is now more frequently defined as the application of machinery to reduce labour-intensive tasks. One of
the upcoming technological trends and beyond is Hyperautomation. In this trend, businesses use a mix of
Artificial Intelligence and Machine Learning to find and automate every imaginable business function. The
analysis of Artificial Intelligence, Intelligent Process Automation, applications of Digital Process
Automation beyond RPA, and the incorporation of automation applications to turn manual work into
automation.
Keywords— Robotic Process Automation, Artificial Intelligence, Intelligent Process Automation, Digital
Process, Automation Applications, Manufacturing Industry, Hyperautomation
PAPER ID : CS639
Music Recommendation Based on Facial Expressions and Mood Detection using CNN Ganne
Vaishnavi, R.Sumathi, Karnatakam Anvitha, Dhathri Bathineed, Balabhadra Nikhitha, Kandregula
Vanaja
[email protected]
ABSTRACT— Music is considered as global language. That can connect the people in all over the world.
Music has the great ability to evoke different emotions in people. Music has great impact on our emotions.
83
ISBN: 979-8-3503-4821-7
It is also a means of mood regulation. In the proposed system, music playlist is created by detecting
emotions. A user’s emotion is detected by people facial expressions. The facial expressions can be detected
lively by system’s camera. The convolutional Neural Network is used for emotion detection. Pygame and
Tinker are used for music recommendation. Feature extraction is performed on input images to detect
emotions such as happy, sad, chill. A music playlist is created based on our mood
Keywords- Universal, regulation, convolutional neural network, pygame, tinker
PAPER ID : CS640
Analyses of Machine Learning Techniques for Sign Language to Text conversion for Speech
Impaired
Jalaparthi Ajay ,R.Sumathi, Arjun K , Durga Hemanth B , Nihal Saneen K
[email protected]
ABSTRACT— Human computer interaction is the research of how individuals and computers interact.
When someone does not understand what we are saying, especially when they do not, hand gestures are an
excellent way to communicate. It is also a fundamental part of human-computer interaction. It's essential
to comprehend hand signals in order to make sure that everyone in the group understands what the person
is trying to say and also that the computer understands what we will be saying. This project's primary
objective is to experiment with various methods for hand gesture recognition. In this project, we use a
camera sensor to identify nonverbal communication. Because most individuals do not really know sign
language because there are not many interpreters, we first tried to create hand gesture recognition. Then,
we built a real-time method for American Sign Language based on deep neural network finger typing,
backed again by an approach with media Pipe. We offer a deep cognitive network (CNN) method for
identifying human hand gestures in photographs taken with a camera. The objective is to separate camera
images from hand motions made during human activity. The training and test data for the CNN were created
using skin model, hand location, and orientation information. The filter is the first thing the hand goes
through before it is classified according to the sort of hand motion it will make. To build this model, we
used computer vision, deep learning, and machine learning. Our Media Pipe model does a good job of
detecting multiple gestures
Keywords— Gesture control, Voice control, Alphabets Sign Language, Convolution Neural Network,
Human Machine Interaction, Human Computer Interaction, Universal Serial Bus, Light Emitting Diodes).
PAPER ID : CS644
SECURING CLOUD IDENTITY INFORMATION UNDER KEY INTRODUCTION
Geetha Priya.M , Kavitha V
[email protected]
ABSTRACT— A fine-grained query approval scheme with trustworthiness take a look at is proposed over
the encrypted spatial facts for location-primarily based totally services. Considering the dispersion of the
spatial facts, a density-primarily based totally area filling curve is supposed to provide the query documents
of the encrypted spatial facts, and inquiry token age and end result take a look at methods are familiar with
make certain fine-grained and apparent spatial query. The proposed scheme empowers the facts owner to
perform finegrained spatial vicinity approval in each the inquiry token age and inquiry end result Trial
effects illustrate that the computational rate of the listing and affirmation shape age methods isn't as a good
deal as that of SHC primarily based totally methods, and the computational and potential expenses of the
84
ISBN: 979-8-3503-4821-7
respectability take a look at technique aren't as a good deal as that of SPR. Moreover, the respectability
affirmation scheme does not gift bogus poor withinside the effects affirmation
PAPER: CS644
Classification of Foods based on Ingredients
Sudipa Guria, Sudarshan Banerjee, Suddhasattwa Bhattacharjee, Swarup Paul, Sobhan Halder,
Priyanka Das
[email protected]
Abstract—Due to the rise of recipe sharing websites, it is now quite easy to find online cooking
instructions that list ingredients and cooking techniques. The development of recipe suggestion systems
has also been a focus of several recipe sharing websites. Little has been done to analyse the relationship
between cuisines and ingredients, while the majority of contemporary food-related research works
concentrate on recipe suggestion systems. The work presented here offers an insight into which cuisines
are the most related and what are a cuisine’s important elements because we know that each cuisine is
unique and has a separate list of ingredients. While the cuisine of food has been regarded as the class label,
the ingredients have been seen as the features. In order to categorise the cuisine of unlabeled data, this work
is focused on selecting the most important ingredients using various feature selection strategies and
classification approaches. This has been done to investigate the fundamental connection between
ingredients and cuisine.
Index Terms—cuisine, ingredients, classification, feature selection
PAPER ID : CS645
Pitch and Timbre Analysis in Acoustics
Aviral Tuteja
[email protected]
ABSTRACT-This project aims to develop a model in deep learning for ‗TOPIC CATEGORIZATION
OF TAMIL NEWS ARTICLES'. The process of learning for regional languages such as Tamil, Hindi
differs from that of English. The task of categorizing Tamil news article by topic is a text classification
task. Most of the applications, such as searching in web, filtering information, recognising language,
readability rating, and analysis of sentiment, already use text categorization. These tasks heavily rely on
neural networks. We did multiclass text classification in this study. Different techniques, ranging from
machine learning to Deep Learning, are used to handle almost all NLP difficulties. Even so, language
localization remains a mystery. For languages other than English, NLP issues are uncertain. Entity
Extraction, prediction, classification, and OCR in sequence modeling are examples of difficulties. Because
the number of individuals utilising local languages like Tamil, Hindi, Telegu in social media network is
growing, it's critical to generate automatic classification process. The goal is to categorize Tamil news
stories into similar subjects (Sports, Cinema,Politics). NB, CNN, SVM algorithms are implemented to
classify the content of Tamil News articles based on the category done in the previous work. The
performance of these algorithms are compared with RNN using Precision, Recall, F1-score are reported in
this study
85
ISBN: 979-8-3503-4821-7
PAPER: CS648
Analysis of Dynamic Scheduling for Edge Cloud Computing
Charan Sai Reddy Vanipenta, Sai Ram Bhukya, Trivikramarka Reddy Koppula, Chandra Shekar
Nelavelli, T.Vignesh, Suneetha Bulla
[email protected]
Abstract— This paper presents a comprehensive analysis of dynamic scheduling for edge cloud
computing. The purpose of this paper is to explore how dynamic scheduling can be utilized to improve the
performance of edge cloud computing systems. First, this paper introduces edge cloud computing, its
applications, and the importance of scheduling. Next, this paper discusses the different types of scheduling
algorithms and how they can be used in edge cloud computing. The paper then investigates the challenges
associated with dynamic scheduling and presents a survey of existing dynamic scheduling solutions.
Finally, this paper highlights the benefits of dynamic scheduling for edge cloud computing and provides
some concluding remarks.
Keywords—Dynamic Scheduling, Cloud computing
PAPER: CS650
An Approach For Cervical and Breast Cancer Classification Using Deep Learning: A
Comprehensive Survey
G S Pradeep Ghantasala, Bui Thanh Hung, Prasun Chakrabarti
[email protected]
Abstract— The most common malignancies in women worldwide are breast & cervical, but very few
researchers have examined how gender expectations affect diagnostic adherence provided sexual
implications and physical contact. Considering these perceptions is essential to enhancing diagnostic
accuracy and services since they may be a major factor in decisionmaking. As cervical and breast treatments
for cancer to be effective, accurate early detection is essential. Machine learning and deep learning are
being used by an expanding population and businesses to analyze vast volumes of data and provide useful
insights. It has become quite frequent in clinical practices to use ML-based techniques to predict the initial
stages of major illnesses like cancer, renal failure, and cardiovascular diseases. Several of the most
prevalent diseases in women include cervical cancer, and early detection could help reduce mortality and
morbidity. The paper provides a comprehensive analysis of the techniques and research issues widely
employed now in the area. The causes, as well as mortality statistics of cancer, are also covered in this. The
goal of the paper is to present a deep learning-centric system for earlier and more accurate breast and
cervical cancer prediction. The Convolutional neural networks (CNN) were utilized in the progress of deep
learning models. This research analyzes the effectiveness of the CNN model employing transfer learning
with a model that has already been trained (VGG16). Considering the results, deep learning algorithms
have the capacity to anticipate disease somewhere at the earliest stage at which fatality rates can be reduced.
Keywords—Cervical cancer, Breast cancer, CNN, VGG16, Classification, Prediction.
PAPER: CS651
A Survey on Exploring the Effectiveness of IOT ased Home Security Systems
Anupriya S.R, Dr. Muthumanikandan V
[email protected]
Abstract—: With the advancement in Internet of things smart homes are rapidly developing. Smart home
is the major key component of Internet of thing. With the help of IOT technology we can stay connected
to our home appliance. Internet of Things is the Associations of inserted advancements that. Contained
physical protests and is utilized to convey and keenness or collaborate with the internal states or the outer
surroundings. Rather than individuals to individuals’ correspondence, IoT accentuation on machine-to-
86
ISBN: 979-8-3503-4821-7
machine correspondence. Smart home connects the physical components of our home with the help of
software and sensors so that we can access them via internet from one place. Building home automation
includes computerizing a home, likewise, mentioned to as a sensible home or smart home. Domestic
machines are an urgent part of the Web of Things whenever they are associated with the web. Controlled
devices are commonly connected to a focal center or entryway through a domestic automation framework.
A smartphone application, tablet PC, personal computer, wallmounted terminals, or even a web interface
that can be gotten to from off-website over the Web are completely utilized by the program to work the
framework. Since all the devices are interconnected and interlinked to one an-another they are lot of
chances for security breach and data theft. If the security layer is easily breakable any third-party attacker
can easily theft the private data of the user. Which leads us to pay more attention to protecting and securing
private data. With the day-to-day development of Smart Home, the safety also got to be developed and
updated day to day the safety challenges of the IoT for a wise home scenario are encountered, and a
comprehensive IoT security management for smart homes has been proposed. This paper acquaints the
status of IoT development, and furthermore contains security issues challenges. Finally, this paper surveys
the Gamble factor, security issues and challenges in every point of view
Keywords-Internet of things, Smart home, protecting private information, security
PAPER: CS654
SOCIAL MEDIA NEWS CLASSIFICATION USING MACHINE LEARNING ALGORITHMS
Dr. M. Jagadeesan, Dr.T.M Saravanan , Dr.P.A. Selvaraj, S. Nithya , K. Snegha ,P. Sowmiya
[email protected]
Abstract— Every day, a great amount of data is generated online in the age of technology. However, an
unprecedented amount of material is being disseminated online, most of it false news. Fake news is created
to draw in viewers, influence people's opinions and actions, raise click through rates, and have a significant
impact on important events like elections. Fake News incorporates material that steers readers in the wrong
directions. It may have unfavorable repercussions in the actual world that seek to deceive people, sway
public opinion, or attract attention or damage reputation. It is necessary to detect fake news mainly for
media outlets to have the capability to attract viewers to their website to generate online advertising
revenue. This project develops a SVM approach and uses this model to detect whether the news are fake
or not. It uses a fake news dataset, which contains News text and corresponding labels (REAL or FAKE).
The pervasive increase in fake news, whether it is produced by humans or algorithms, has a severe social
and political impact on society and individuals. The rapid turnover of news in the age of social networks
makes it difficult to assess its veracity in a timely manner. Tools for automatically detecting bogus news
become essential. K-Nearest Neighbor and a classification based on a support vector machine are utilized
in this case to overcome the aforementioned issue. In contrast to other research that merely took particular
sentences or phrases into account, The purpose of this study is to investigate a false news posture detection
mechanism based on the title and text of the news. Python 3.7 is used in the project's development.
Keywords—Machine Learning; K-Nearest Neighbor; Support Vector Machine
PAPER: CS655
ANIMALS DETECTION SYSTEM IN THE FARM AREA USING IOT
Ms.K.Chitra, Dr.A.Tamilarasi, Dr.M.Pyingkodi, K. Nanthini, Ms.V.Sneka, Ms.S.Swetha,
Ms.P.Vishalini
[email protected]
Abstract -The project's primary objective is to protect the crops from being ravaged many times by wild
animals like elephants ,wild pigs,monkeys.there will be a large amount of loss to farmers.To avoid the
financial losses for the farmer, so that it is very much important to protect agricultural fields or farms from
the animals. Farmers cannot block entire fields or remain on the field all day to secure it. to safe-guard the
87
ISBN: 979-8-3503-4821-7
agriculture field and intimate the farmer via call alert when any animal is detected. This project has Arduino
Uno used to control the Ultrasonic sensor while both are connected to each signal transmitted to the GSM
module to transmit a call notification and alarm sound.
Keywords — Arduino uno , Ultrasonic sensor ,Global System for Mobile communication)GSM module
PAPER: CS657
Extensive Review on Predicting Heart Disease Using Machine Learning and Deep Learning
Techniques
S.Hemalatha , Dr.T.Kavitha, K.Chitra, B.Babitha, N.Kaya, R.Keshika
[email protected]
Abstract— A variety of illness that affect your heart are referred to as heart disease. The accurate diagnosis
of heart disease can reduce the risk of serious health problems, whereas an inaccurate diagnosis can be
fatal. This research compares the findings and analyses of the cardiovascular illness using various machine
learning and deep learning methods.
Keywords— Cardio Vascular Disease; Machine Learning; Deep learning
PAPER: CS658
An effective architecture for Image colorization using Adaptive GAN
Ravindhar N V, K Akshaya, Pokala Ramadevi, Madithati Yuvateja Reddy
[email protected]
Abstract— More attention has recently been paid to image colorization. In order to produce realistic
results, traditional reference-based methods frequently rely on external colour pictures. An extensive
picture database or internet search engine will be required to obtain such instances. At a reasonable cost,
recent deep-learning-based algorithms may colorize photos automatically. On the other hand, inconsistent
hues and poor artifacts are always present. Utilizing the extensive and diverse colour priors stored within
a pre-trained Adaptive Generative Adversarial Network (GAN), our goal in this study is to restore vibrant
colors. In particular, we incorporate matched features comparable to exemplars into the recoloring
procedure by using feature modulations and a GAN encoder to "extract" them. Our system could generate
vivid colors with just one forward pass thanks to the dynamic generative colour prior and sensitive designs.
Additionally, changing the GAN hidden codes makes it extremely simple to obtain a variety of outcomes.
By traversing GAN latent space, our method enabled controlled and smooth transitions while maintaining
the advantages of Adaptive GAN interpretable controls. Our approach outperforms previous efforts in
terms of execution, as evidenced by extensive testing and client research.
Keywords—Image processing, Image enhancement, Colour Corrections, Image features, Deep learning
PAPER: CS659
Recognition and Classification of Image Based Coal and Coal Gangue Using Deep Learning
Techniques
T Kalpana, R Thamilselvan, K Chitra, M Pyingkodi, B Dakshinya, N RManu, SGokul
[email protected]
Abstract - Observing coal features is the first step in learning about coal. Here, a coal gangue detection
system was created that enables users to do a search even when they are unsure of the coal's name by
looking at specific coal features. Currently, machine vision is used to extract and analyze color, size, shape,
and surface texture in coal classification. Even if the new extraction margin method can be applied
generally, the margin of the extracted polygon's form and the edge of the real image's shape are still not
satisfied. Finding the gangue in the coal is the project's goal. Then, depending on the number of pixels
representing the gangue colors, the entire gangue level in the coal data is determined and shown. This helps
determine the coal's quality. Even difficult-to-quantify qualities, like coal gangue, could be quantified in
88
ISBN: 979-8-3503-4821-7
the future by academics who broaden their focus to include other features. The coal dataset is classified
using an artificial neural network.
Keywords - Coal Classification, Centroid Contour, Deep Learning, Neural Network.
PAPER : CS660
Conversion of ECG signals and heart disease detection using Machine Learning Algorithm
Dr. T. Kavitha , S.Hemalatha , Ashiq Mohamed S , EsakkiRaja M , Kowshick J
[email protected]
ABSTRACT - Machine learning is utilized in a wide range of global sectors. In the healthcare industry,
there are no exceptions. Machine learning may be very helpful in predicting the presence of cardiac
problems, anomalies of the locomotor system, and other disorders. Such knowledge, if expected
beforehand, can provide clinicians with critical information, allowing them to tailor their diagnostic and
treatment approach for each patient. In order to foresee potential human cardiac abnormalities, we are
investigating machine learning techniques. We present a DNN model made for this purpose using data
from the code (Clinical Outcomes in Digital Electro cardiology) study. We use it to categorise six different
ECG anomalies that are believed to be indicative of both rhythmic and nonrhythmic diseases.
Keywords - Neural Network; ECG; DNN; EKG
PAPER : CS661
Forecasting the Solar Power with Differentiation of Data Figures using Neural Networks
Dr. T. Kavitha , S.Hemalatha , R.V.Naveen kumar , S.Sakthi Balan , V.Sharan
[email protected]
ABSTRACT - The industrial sector relies heavily on solar power forecasting because it is dependable and
readily available. High solar energy penetration is essential for the financial sustainability of power systems.
The fact that solar energy is actually renewable is the most significant benefit of all the features of solar
panels. Forecast models incorporate information from numerical weather prediction (NWP) factors such as
cloud cover, sunshine, radiation, precipitation, etc. However, there is NWP Uncertainty in the assumption
of the atmosphere's starting state. Advanced predictions are made in the meantime by machine learning
algorithms. We need to estimate and compare the accuracy rates between Ann algorithm and Rnn algorithm
for the outgoing solar power in this study using the dataset from the previous month.In the upcoming month,
the crucial machine learning algorithm will be created. The most extensively used "multi-layer Neural
Network" model of ANN and RNN should anticipate and compare electricity costs. Our research aims to
accurately forecast solar power. The technique of forecasting the solar energy collected from solar panels
for future use is attained by machine learning algorithms. The prediction and comparison will be carried out
using the Artificial Neural Network and Recurrent Neural Network concepts in Machine Learning.
Keywords -NWP : Long short-term memory, recurrent neural network, artificial neural network, feed
forward propagation, and backpropagation
89
ISBN: 979-8-3503-4821-7
PAPER : CS662
MALWARE PREDICTION AND CLASSIFICATION FOR ANDROID APPLICATIONS USING
MACHINE LEARNING TECHNIQUES
T Kalpana , R Thamilselvan, D Sivabalaselvamani, M Pyingkodi, S V Gururohith, A Jayasarathy, P
K Kamalesh
[email protected]
ABSTRACT - Numerous paid and open-source applications have been produced as a result of the
increased use of Mobile phones. These programs enable users to perform a wide range of tasks, including
as communicating, playing games, and carrying out financial and educational tasks. Malicious app growth
has become a significant problem that is holding back the development of a healthy mobile ecosystem.
According to recent studies, new dangerous Android apps are released every 10 seconds. Users of the
Android operating system can download and install software from unreliable websites including file-
sharing websites and third-party app stores. The issue of malware infiltration has become so serious and
significant that, according to recent statistics, 98% of mobile malware targeted Android smart phones. In
this project we using SVM, KNN and CNN algorithms.
PAPER : CS664
MULTICLASS SKIN CANCER IMAGE CLASSIFICATION USING CONVOLUTIONAL
NEURAL NETWORK
Ramya J , Vijaylakshmi H C , Huda Mirza Saifuddin
[email protected]
ABSTRACT - Skin cancer is the most dangerous, having one-third diagnosed rate in worldwide. It is
supposed to spread vastly to other body parts when it is not detected at beginning stage. Skin cancer
diagnosing system contains three major steps such as segmentation, feature extraction and classification.
Firstly, images are pre-processed using 2D median filter to remove some artifacts like thin hair, gel, air
bubble etc. followed by segmentation using discrete wavelet transform to extract lesion region from
background skin region is described in our last paper. In this paper, we have passed previously segmented
images to Convolutional Neural Network (CNN) for feature learning and classification. Experimentation
conducted on three classes of PH2 dataset, classification results are satisfactory and accurate.
Keywords - convolutional Neural Network, Features, Classification, PH2
PAPER : CS665
Analysis Of Irregular Menstrual Cycle And Its Consequences
Sneha M , Sai Sravani K , Nithya S , Dr.R Kaladevi
[email protected]
ABSTRACT - Women in the 21st century is diagnosed with many kinds of health issues which are later
found to be serious. One of the most worrying diseases today is Polycystic Ovarian Syndrome(PCOS)
which is highly dangerous to an extent of affecting women's reproductive life to a huge extent. This causes
various disorders like infertility, endometrial cancer, etc. The most prevalent symptoms of this disease are
oily skin, acne, irregular menses, anxiety disorders, and hypertension. These women are also at increased
risk for mood disorders, such as depression and anxiety, and non-alcoholic fatty liver. An estimation states
that one in five women suffers from PCOS[1]. Many women wake up and notice the symptoms only at the
90
ISBN: 979-8-3503-4821-7
time of conceiving. This is happening because of the lack of knowledge about this disorder. Our aim is to
develop a medium, which helps the end user (Specially made for women) to know their Health level
according to their period cycle. Data prediction provides the end user with Suggestions based on their input
by classifying the data provided by the user using machine learning algorithms. In this way, the tremendous
increase could be controlled and diagnosed on time.
Keywords - PCOS, Image Analysis, Menstrual Cycle, Machine Learning, Expert System.
PAPER : CS666
ACMapping: Agricultural Crop Mapping/Retrieval Using Synthetic Aperture Radar (SAR) Data
N. Sravani , R. Mitravinda , P. Ramesh Kumar , N. Neelima , K L Sailaja
[email protected]
ABSTRACT - Agriculture appears to play a significant role in both a nation's food security and economic
development. Image processing techniques can be applied to Synthetic Aperture Radar (SAR) images with
high spatial resolution. These images are backscattered to overcome the noises like clouds, smoke, rain,
trees, etc. Traditionally Optical images are used for crop monitoring which has less brightness of pixel as
it depends only on one variable, the amount of light reflected by the earth’s surface. Whereas in SAR
images the brightness of each pixel primarily depends on at least 8 variables which give high resolution
than optical images. The main aim of our project is to develop a model in Agriculture Field for Crop Growth
Monitoring using SAR data. We used Sentinel 2 imagery data and processed them with Advanced Remote
Sensing techniques. For crop monitoring analytics, there are different types of Vegetation Indices available.
These indices help in remote sensing by providing useful insights of crops. In general, the Normalized
Difference Vegetation Index (NDVI) is the most widely Vegetation Index to report biomass, density due
to its versatility and credibility and it gives better accuracy during the whole crop production time. Using
this parameter, we develop a methodology for Agricultural Crop Mapping/Retrieval using SAR images.
Keywords - Synthetic Aperture Radar, Vegetation Index, Remote Sensing, Crop Growth, Sentinel 2.
PAPER : CS668
A Survey On Text Summarization in Urdu Language using Machine Learning Techniques
Prathi Himaja , Guduru Anvitha , Syed Fasih Ahsan , Manikya Chowdary , T. Vignesh
[email protected]
ABSTRACT - The principal commitment of this paper is to analyze and understand the various text
summarization methods. We initially present a few ideas related to the extraction of particular text from
the. Specifically, a few difficulties in extraction of summarization of text and various methods presented.
Additionally, this paper concentrates on Urdu Language summarization and issues which can be faced by
Urdu Language summarization. The challenges examined are nonexclusive and relevant to each
conceivable situation in text summarization analysis. Then, existing technology of extractive synopsis
procedures are examined that attention on the distinguished difficulties. This paper presents a great
understanding for text summarization techniques summarization there since 40 years. This method gives
the best solution to the problem by suppressing the content and giving important summary of algorithm is
easy to applicable. In this method it removes the extra sentences which does not make any sense which
improves informative content.
Keywords - Extraction based summary, Natural Language Processing, Urdu language
91
ISBN: 979-8-3503-4821-7
PAPER : CS675
EFFICIENT ROAD STRUCTURAL DESIGN AND TRAFFIC ACCIDENT ANALYSIS USING
SUPERVISED LEARNING ALGORITHMS
Gonuguntla Hruthik , Siddhartha Shankar Swain , Ronanki Tharun , Srithar S , Seethala Chiranjivi
Sai Vara Prasad , Aravinth S S
[email protected]
ABSTRACT - The flow of traffic analysis can be helpful for to improving the traffic quality and protect it
from the conditions to cause danger, risk, or injury. Our goal is about reducing the accidents risk which
involves designing the roads and principles. The previous road designing patterns of highly accidental risk.
Especially some of the circumstances of accidents may be related to the driver such as driver fault, health
problems, or also vehicles such as brake failure, steering, or the road itself such as poor sidewalks, and lack
of road equipment, and safety measures. In this paper, we talk about some of the best road designing
techniques in the world. ‘Double crossover diamond interchange (DCD)’, is a true way to go more people
reduce the vehicle striking violently against another. One more design is that a ‘roundabout’, is a circular
common intersection where the vehicles can travel anticlockwise around that center part. In This Modern
roundabout, there is no need for traffic signals or speed breakers. When the vehicles enter the roundabout
zone, if traffic is there in the roundabout, then go to the circulating roadway area and go to their desired
street. Find the exact models to predict the deaths and severe injuries caused by road accidents. This
analysis needs to classify the various models of machine learning and data science algorithms for instance
Linear Regression, Decision Tree, Random Forest (RF), and Naive Bayes Classification and Support
Vector Machine (SVM). The decision tree algorithm has a root node and contain child nodes. But Random
Forest is having multiple decision trees. Random Forest has the best performance with 79% accuracy then
decision tree with 70% accuracy, SVM with 75%, Logic Regression with 72%, and Naive Bayes with 73%
accuracy.
Keywords - Double crossover, Decision Tree, Diamond interchange, Support Vector Machine,
Roundabout, Naive Bayes Classification.
PAPER : CS678
AGROCULTURE WEB APPLICATION
Dr.Sumithra.A , Sangeetha.M , Sneha.R , Suganya Devi.M

[email protected]
ABSTRACT - Agriculture is the backbone of our country. Mostly 60-70% of people in India depends
either directly or indirectly on Agriculture for food for their own living. As India is also an agriculture
dependent country , it is the major platform for India's development. Eventhough, Agriculture is the main
occupation of India ,farmers have both advantages and disadvantages too. Farmers usually invest most of
their money on agriculture. But they were not correctly rewarded to their hardwork. They only get less
amount of money compared to the money they invested on it. Ss ome farmers who were new to agriculture
may also don't know much about agriculture like which pesticides need to be used on certain crops, how
many times the crops should be irrigated per month, how many months it will take for cultivation or maybe
how it should be cultivated. So, the vital idea of our project is to create an web application which will help
the farmers to solve these problems for selling their goods as it is a trust-worthy and beneficial way for
farmers. It would be profitable for both the buyer's side and the seller's side.
Keywords - Ordering , Time management , Online products
92
ISBN: 979-8-3503-4821-7
PAPER : CS680
A comparative analysis between single & dual-handed Bangladeshi Sign Language detection
using CNN based approach
Gourob Saha Surjo , Biplob Kumer Ghosh , Mohammad Jahangir Alam , Mahamudullah , Md Razib
Saurabh Bilgaiyan
[email protected]
ABSTRACT - In order to help the vast majority of the population learn sign language, specialists today
frequently apply machine learning techniques. This project seeks to create a model that can recognize the
letters in Bangladeshi Sign Language (BdSL) using a deep learning approach. This study is a comparison
between the recognition of single-handed and dual-handed Bangladeshi Sign Language. Dataset, KU-BdSL
is selected to train the single-handed BdSL and we selected dataset - BDSL 49 for dual-handed BdSL. We
suggested pre-trained models of CNN based approach for the purpose of detection and classification. 30
different alphabets from single-handed BdSL and 36 different alphabets from dual-handed BdSL could be
classified and recognized by our CNN models. Three pre-trained CNN models were employed. VGG16
outperformed the others by a wide margin. It correctly identified single-handed gestures with 98%
accuracy. Regarding the ability to recognize dual-handed Bangladeshi Sign Language, VGG16 once again
outperformed competitors with an accuracy of 90%.
Keywords - Bangladeshi Sign Language, Convolution Neural Network, CNN models, Hand gestures.
PAPER : CS684
Big data analytics and machine learning techniques to manage the smart grid
M.Elavarasi , Dr. Mahesh Kotha , Ramakrishna Kolikipogu , M. V. B. T.Santhi
[email protected]
ABSTRACT - Big data is a crucial part of the Energy Internet. Integrating renewables and smart grids is
vital and exciting. Any power plant must collect and evaluate data to make educated choices. This study
examines big data analytics in renewable energy based power plants. This paradigm implements big data
analytics for renewable energy utilities and smart grids. The authors describe a five-step machine-learning
approach to predict smart grid dependability. Using 64,000 occurrences and 12 attributes from an intelligent
grid data system, we expected the system's stability using three machine-learning algorithms. The penalized
linear regression model achieves 95% accuracy using 70% of the training data. The random forest model
is 84% accurate vs 77% for the decision tree. CNN and gradient-boosted decision tree models obtained
86% classification accuracy. This study's small dataset prevents extensive data analyses. Cloud computing
and real-time event analysis are well-suited for a data analytics infrastructure. Future research should
include data from additional countries and renewable energy sources.
Keywords - Big data analytics, smart grid, machine learning, Decision tree.
93
ISBN: 979-8-3503-4821-7
PAPER : CS688
DRUG CONSUMPTION DETECTION BY EYE USING IMAGE PROCESSING TECHNIQUES
Pyingkodi M , Thenmozhi K , Wilfred Blessing N.R , Archana K , Arun Hiruthik X , Dinesh Kumar
V
[email protected]
ABSTRACT - Drug usage is a significant global economic issue that results in countless losses. This
study suggests a convolutional neural network-based image processing method for identifying drugged
eyes. The front-end packages already in use for this article collect information from eye images using the
CNN algorithm. However, it can take some time. As a result, the suggested approach may be used to swiftly
and automatically identify drugged eyes. The acquisition of the input image, picture pre-processing, and
image processing are the main processes in the suggested technique.locating the reddish regions,
emphasising the afflicted regions, confirming the training set, and presenting the findings. Some eyes may
go unnoticed, such those that have been socially poisoned. This technique was tried on drugged eyes at
different phases, both with and without drug consumption. The technique was used to find a white region
of the eye in a given input picture. Images of both drugged and unaffected eyes were presented for training
purposes. In order to choose the optimum colour model for this strategy, photos were first transformed to
colour models. The Support Erosion technique and the Local Binary Pattern were both used in the model
construction for feature extraction. This technique can detect drugged eyes with an average accuracy of
95%.
PAPER : CS689
SMART MANUFACTURING IN INDUSTRIES USING INTERNET OF THINGS
Lakshmi likhith yadla , Raghavendra Avinash Mudragada , T.Vignesh , Harikrishna poka
[email protected]
ABSTRACT - IIoT plays very important roles in the field of automation in mechanical/mechatronics
industry. IIot is the sub part of the Internet of Things (IoT). The most significant advantage of IIoT is that
it allows manufacturers to automate and thus optimize their operating performance. Various sensors can be
used in the fields of IoT/IIoT which can be used in the production industries to monitor the work
performance of various processes of automation process, gathering the data to establish normal difference
operation of find out even minor variations in that performance. Sensors automatically record data such as
energy consumption, temperature, humidity, operating time, maintenance, and outputs from production
lines. Weather instruments, aircraft, vehicles, and any other machinery with pressure functionality. Pressure
sensors in systems can be used to measure other variables such as fluid/gas flow, speed, water level, and
altitude. Sensors are used to detect smoke, measure temperature, and regulate pressure, among other things.
Because analogue signals are continuous, even minor changes in physical variables such as temperature or
pressure can be accounted for. In the paper, we have summarized the advantage and disadvantages of IIoT
and feature scope of IIOT in detail.
Keywords - Industry internet of things, pressure sensors, smart manufacturing, IIoT Applications,
Augmented reality.
94
ISBN: 979-8-3503-4821-7
PAPER : CS690
Asthma Disease Risk Prediction Using Machine Learning Techniques
Pyingkodi M , Thenmozhi K , Wilfred Blessing N R , Selvaraj P A ,Kamdev Kumar ,Aadarsh V

,Mariya Arockiya Akash A
[email protected]
ABSTRACT - The signs of asthma, a long-term inflammatory condition of the airways, include wheezing,
throat tightness, coughing, and breathing difficulties. The attack of an asthma, which can be fatal, is the
fast worsening of these symptoms. Another severe, irreversible airflow restriction in the lungs is caused by
respiratory COPD, which encompasses emphysema and chronic bronchitis. In this project, a machine
learning-based algorithm is for predicting asthma risk is presented (ML). PEFR, which are widely used
external tools like peak flow meters and recognized asthma risk predictors, are frequently monitored. This
study shows a relationship between the ambient particle matter(PM) and the weather outdoors. According
to the best peak flow value each person was able to acquire, the results are divided into two groups: Safe
and Risk circumstances. The link between indoor PM and weather data is mapped to the found values using
a convolutional neural network (CNN) architecture. The suggested method's root mean square and mean
absolute error accuracy metrics are contrasted with those of current deep neural network (DNN)-based
methods. Additionally, the accuracy of the classification methods KNN and SVM are carried out. The new
data set's asthma category may be predicted more accurately thanks to the application of SVM, KNN, and
CNN classification. Python 3.7 is the coding language employed.
Keywords - Asthma Disease, KNN, CNN, COPD, Particle Matter, Support Vector Machine
PAPER : CS691
A Structured Analysis on IPL 2022 matches by approaching various Data Visualization and
Analytics
Selva Birunda S , Nagaraj P , Jency A Jebamani B , Revathi B , Muneeswaran V
[email protected]
ABSTRACT - The main purpose of making this report is to give the information to the franchise to make
their decision on buying the players by bidding in the auction based on their performances. There are many
models of reports which describe the IPL seasons. This report's purpose is to give a brief analysis of the
IPL 2022 15th league. In this report, much of the data from the datasets of matches is utilized and analysis
is made by using Python packages specifically, Pandas, NumPy, and matplotlib. By making graphs and
tables a brief analysis with complete knowledge of an IPL season is given. This analysis describes every
good performance and gives good chance to the best players.
Keywords - franchise; emerging player; broadcast; tournament; statistics
PAPER : CS692
DEVELOPMENT OF TEXT CHARACTERIZATION USING NATURAL LANGUAGE
PROCESSING TECHNIQUES
Nagaraj P , Muneeswaran. V , Karthik Goud B , Vigneshwar Reddy G , Arjun k ,Nagaraju B
[email protected]
ABSTRACT - Purpose : Text summarization means it reduces the text without losing the content. By
this, we can reduce the text size. Method : In this work, we attempt to use tokenization and stop word
95
ISBN: 979-8-3503-4821-7
removal to extract the useful and main content of the text. The model read the given text and performs
tokenization, it considers the punctuations, and words as the token and it will eliminate the white spaces,
then it removes the stop words. We calculate the cost of the sentence in the authentic corpus. By this, we
can obtain our minimized text. Result : The proposed system can work with an accuracy of 96%, which
means the summarized text is accurate to the given text.
Keywords - tokenization and stop word removal to extract the useful
PAPER : CS697
Secure Examination System with YOLOV3 & Dlib
Gade Bala Rohith Reddy , R. Aishwarya , R. Yogitha
[email protected]
ABSTRACT - COVID-19 has aided in the advancement of remote learning. Students continued their
studies despite the closure of schools and institutions by using tools such as Microsoft Teams, Zoom, and
a variety of other video conference software. Exams, on the other hand, have yet to find a remedy. Some
have merely canceled them, whereas others have altered them to assignment forms that students can cheat
from online sources just by searching for the answers. There should be an answer if the manner we tend to
live is to become the new standard. Online tests push students to take exams from the desired location,
During the entire exam, they will be watched and monitored. Now we will create a system in Python that
can capture or detect malpractice moments during the exam which includes various features and manage
more students at a time. Here, we develop a system that can avoid physical monitoring by the teacher. The
developed multi-model AI can withstand any number of participants without the need for human
intervention. During the test, We record video using the built-in camera and captured windows through the
exam. The test taker’s face is identified, as are eye-ball moments, and their emotions are predicted. His
feature points are identified to determine his head position. Among the different objects that can detect a
person in the background and objects like a book or a pen. By combining these models, an intelligent rule-
based inference system is created that can determine whether examination fraud occurred. Regarding object
detection, the YOLOv3 object identification method is a superior choice for this project. Exams' main
objectives are to assess students' abilities and promote more focused learning. Exam results have the
potential to greatly increase a candidate's possibilities.
Keywords - YOLO, Ai, Intelligent rule-based, Face detection
PAPER : CS703
Hate Speech Analysis using Supervised Machine Learning Techniques
Pyingkodi M , Thenmozhi K , Chitra K , Karthikeyan M , Pradeep M , Suriya Prakash T D ,
Viswanathan T J
[email protected]
ABSTRACT - Hate speech is a serious issue that can have harmful consequences for individuals and
groups. In an effort to combat this issue, automated methods for identifying and categorising hate speech
in text and other media have been developed using machine learning and deep learning techniques. These
methods involve training machine learning algorithms on large datasets of labelled text, audio or visual
data, and then using these algorithms to identify and classify new, unseen examples of hate speech. Machine
learning approaches can be effective at detecting and classifying hate speech, but they also have limitations
and can be prone to errors, particularly if the training data is biased or if the algorithms are not carefully
designed and evaluated. It is important to thoughtfully examine the ethical and social consequences of using
machine learning to identify hate speech and to ensure that these methods are transparent, fair, and
accountable.
Keywords - Hate Speech, Emoticons, Random forest, Naive Bayes, Logistic Regression, Support Vector
Machin
96
ISBN: 979-8-3503-4821-7
PAPER : CS704
Data Analytics for Phishing Attack Detection using Deep Learning Techniques
M.Karthikeyan , Dr.S.Vijayachitra , Dr.M.Pyingkodi , T.Aadhish , P.Dhayananth , G.Goodwin
[email protected]
ABSTRACT - As the process of digital transformation continues to advance, the challenges to global
cyber security have increased. As a result of digitalization, hackers have more alternatives. Phishing efforts
are the initial type of cyber-attacks, and they aim to get individual user passwords. Phishing is a common
tactic used by hackers to trick people into giving them access to a company's networks and digital assets.
Phishing is a type of cyber-attack in which a victim is tricked into divulging sensitive information including
account passwords, bank account details, ATM pin numbers, and other personal information. Phishing-
style cyber-attacks hurt individuals by deceiving them into disclosing important information. In this paper,
the phishing attack dataset will be taught and evaluated to see which deep learning method yields the best
results.
Keywords - Artificial intelligence, machine learning, deep learning, recurrent neural network, multi-layer
perceptron, phishing assault, and phishing attack detection.
PAPER : CS705
INFORMATION EXTRACTION FROM TEXT MESSAGES USING NATURAL LANGUAGE
PROCESSING
Bapanapalli Harshitha Sai Durga , Nerella Venkata Rama , Kalakota Sai Sanjana ,Radha Mothukuri
, Yusuf Baig,T.Vignesh
[email protected]
ABSTRACT - In unstructured text, process of looking for patterns is called text mining. Finding particular
elements in documents written in natural language is a related task of information extraction (IE). Text
mining has grown as an importance in study field. It is discovery of the new and past undiscovered
information by computers from a variety of hand-written sources automatically. We currently live in a time
marked by extreme stress and mental illnesses. A greater number of people are committing suicide as a
result of the excessive levels of stress and strain because they tend to increase the amount of people who
exhibit risky inclinations. Family conflict, work dissatisfaction, health issues, etc. are common sources of
stress. In the modern computing environment, people are pleased to express their opinions and sentiments
on social media with friends and family members by using tools like electronic communication. The
suitable text mining technique can increase efficiency and cut down on the time and effort needed to retrieve
relevant information. The Natural Language Processing technique which is used to extract important
information from text messages is briefly discussed and analyzed in this work.
Keywords - Text Mining, Information Extraction, Text Messages, NLP, Patterns.
PAPER : CS707
A SURVEY PAPER ON WEED IDENTIFICATION USING DEEP LEARNING TECHNIQUES
U. Gayathri , Dr. V. Praveena
ABSTRACT - Agriculture is the fundamental art every country will embrace for its economy, which is
necessary for all living things to live. The Farmers can reduce their total workload by applying technology
to improve grain quality. The genuine steps that need to be taken by the farmers' weed control and
eradication practices result in increased plant growth and improved water-use efficiency, thereby allowing
them to maximize crop yields with less effort and a higher percentage of growth. An Automatic detection
and classification of weeds can play an important role in weed management to maximize the use of
97
ISBN: 979-8-3503-4821-7
resources and minimize the cost of weed control. It is critical to use deep learning technology to identify
and estimate weed. This survey paper describes the progress in weed detection and classification using the
most recent technologies in the field of artificial intelligence, and image processing techniques. It includes
detailed literature on the major strategies.
Keywords - Agriculture, Weed identification, Deep learning, Image processing, Crop production, SVM
algorithm, Random forest, etc.,
PAPER : CS711
A novel approach for monitoring agricultural production process using wireless sensor networks and
Machine Learning
Subhashree Rath , Aditya Arabale , Senthil Kumar R , Charanraj K R , Nishanth S B
[email protected]
ABSTRACT - The majority of nations rely heavily on agriculture. In India, agriculture directly supports
more than half of the country's population. The yield of a given crop is influenced by a number of variables,
including the climate, wind speed, soil quality, humidity, etc. The growth of a crop is impacted by these
components' ongoing variability. The agricultural industry has benefited from technological advancement.
In the agriculture industry, wireless sensor networks and crop yield prediction have had a significant
impact. This study introduces a revolutionary Precision Farming method and demonstrates how sensor data
can be effectively utilised. The Crop Yield Prediction model receives real-time input from the sensor-
generated data. This strategy aids us in getting more accurate results. The research suggests a mobile
application that, every split second, sends data from wireless sensors used in precision farming as an input
to a crop production forecast model. Additionally, these sensed parameters from different users can be used
as a training dataset. This method not only makes the most of the sensor data but also predicts crop yield
with accuracy and promptness.
Keywords - Precision Farming, WSN, IoT, Machine Learning
PAPER : CS715
Performance Evaluation of Transfer Learning Based Models On Skin Disease Classification
R. C. Suganthe , N.Shanthi , M.Geetha , R.Manjunath , S.Mithilesh Krishna , P.Muthu Balaji
[email protected]
ABSTRACT - Skin is a vital organ in our body and it even serves as a shield against harmful rays. If skin
gets some of the problems, then it would affect the whole body. Major reason for the skin cancer is due to
ultraviolet rays, exposing body to chemical and genetical problems. It can spread to the body's inner cells
and cause DNA damage. Diagnosing skin cancer at earlier stage becomes more important in order to
decrease the malignancy of the disease. Deep learning technology can solve the complexity and helps us
to diagnose skin cancer. Deep learning is a part of machine learning where it can work like human brain
and can work with its past experience of solving complex problems. Convolutional neural networks can
classify skin lesions with much higher accuracy so in this experiment, deep learning model is used.
Convolutional neural networks, which may have up to 30 layers, are designed exclusively for image
analysis. Categorization, image processing, and segmentation are some of the applications. In this study,
transfer learning models including VGG16, VGG19, Inception V3, Convolutional neural networks, and
ResNet were used. Benefits of CNN is the ability to develop internal representation of 2D image and it can
learn more than SVM, KNN etc which becomes key point in using CNN.
Keywords - Skin cancer, Classification, CNN, VGG16, VGG19, ResNet, INCEPTION V3.
98
ISBN: 979-8-3503-4821-7
PAPER : CS716
TEXT ANALYSIS ON TWITTER DATA USING LSA AND LDA
T.Karthik Venkat Sai , K.Lohith , M.Padma Sai , K.Tejaswi , P.M.Ashok Kumar , Karthikeyan C
[email protected]
ABSTRACT - The news ecosystem has changed in the modern years from outdated print media to social
media sites. Because social media platforms enable us to absorb news much more quickly and with less
restrictive editing, fake news is disseminated at an astonishing rate and scale. More and increasingly
individuals are using social media as the world becomes more digital since it makes connecting with others
relatively simple. But false information is misguiding people. Although fake news is simple to propagate,
its effects can be disastrous. Bogus news has frequently resulted in uncontrolled circumstances that killed
numerous people. People with limited education might quickly become acclimated to bogus news. Instead
of checking the accuracy of the information, they accept what is provided to them. This can be overcome
by using Text mining, Statics concepts which can detect fake news more precisely when compared to
several machine learning methods. This proposed system is on the analysis of fake news and detection of
hatred news using t-distributed Stochastic Neighbor Embedding(t-SNE) to check dimensionality reduction
and topic modeling using Latent Dirichlet Allocation(LDA) and data pre-processing using Latent Semantic
Analysis (LSA).
Keywords - Fake News, Twitter, social media, t-sne, LSA, LDA
PAPER : CS717
Efficient load balancing for Micro Services based applications
Javahar A , Dharun R , Ananth R , Arun Ritthik K K
[email protected]
ABSTRACT - Any micro service-based programme, including alumni management software, must
include load balancing. The effective use of resources is made possible by load balancing, which also
ensures that application performance is unaffected by unexpected increases in requests or user usage.. In
an alumni management application, load balancing can be achieved with a combination of techniques such
as round-robin scheduling, weighted load balancing, and dynamic scheduling. These techniques help to
ensure that the application remains up and running, even during peak usage times. Additionally, load
balancing can be used to ensure users get the best possible experience when using the application. By
distributing traffic among multiple back-end services, users can expect faster response times and more
reliable performance.Load balancing in a micro services-based alumni management application is a way
of distributing incoming web traffic across multiple back end servers in order to efficiently manage traffic,
reduce latency, and improve overall performance. Load balancing works by distributing incoming requests
across a cluster of servers, and can be done either statically or dynamically. Static load balancing assigns
each request to a specific server, while dynamic load balancing uses algorithms to decide which server
should handle each request. In the alumni management application, load balancing is important for ensuring
that requests are handled efficiently and that the system is not overwhelmed by too many requests. This is
especially important when multiple users are accessing the application simultaneously. By distributing
requests across multiple servers, the application can remain responsive and reliable even under heavy load.
Keywords - Load Balancing, Micro service, Alumni Management, Round Robin, Response Time,
Scalability
99
ISBN: 979-8-3503-4821-7
PAPER : CS726
Cross Site Scripting (XSS) vulnerability detection using Machine Learning and Statistical Analysis
HARISH KUMAR J , Dr. J GODWIN PONSAM J
[email protected]
ABSTRACT - In our current technological development, usage of social networking, e-commerce, media,
and management, web application plays a very indispensable role on the Internet. Organizations use web
applications to reach information to the public, e-commerce sites like Amazon and Flipkart use web
applications to sell their products, and social networking sites like Facebook and Instagram use web
applications. Many other services are provided on the web. Every mobile application will have its
equivalent web application. Web Application Security plays a very vital role around the world. Cross Site
Scripting (XSS) attacks are by far the most common and widely used method for stealing data from web
applications. This paper discusses the XSS vulnerability detection using different deep learning and
machine learning models. XSS attacks are a common type of web-based attack in which malicious code is
injected into a website or web application, allowing attackers to steal sensitive information or perform other
malicious actions. To ensure web-based systems' security, XSS attack detection and prevention are
essential. If the attacker successfully executes the XSS script, then the website will be compromised, and
the attacker can steal sensitive data. The Open Web Application Security Project (OWASP) has listed XSS
attacks as a top three risk to web applications. This research paper proposes a novel approach for detecting
XSS attacks using different models. Deep learning algorithms such as Long Short Term Memory (LSTM),
Convolution Neural Networks (CNN) and boosting algorithms such as AdaBoost and Gradient Boosting
algorithms, and classification algorithms such as Logistic Regression (LR), Support Vector Machine
(SVM), K-Nearest Neighbour (KNN), Random Forest (RF), Naïve Bayes (NB), and Decision Tree (DT)
algorithm for the detection of XSS attacks. To evaluate the effectiveness of our approach, we conducted
experiments on a dataset of real-world XSS attacks and non-attack web requests. Our experiments showed
that our machine-learning model was able to accurately identify XSS attacks with a high degree of
accuracy, outperforming several baseline approaches. Overall, our research demonstrates the potential for
using machine learning to detect XSS attacks effectively.
Keywords - Attack detection, machine learning, Web Application Security, XSS attacks
PAPER : CS731
A SURVEY ON ANALYZING SOCIAL MEDIA DATA AND THE DATA FROM PAST FOR
EMERGENCY MANAGEMENT
Gugan I , Kavitha M S , Karthikeyan N , Karthik S
[email protected]
ABSTRACT - Disaster management is the need of the hour in countries like India, where Information
science in still in its developing stage. In order to address this issue, a survey on works related to disaster
management in the social media front by various researches is done in this work. The works considered for
the survey include different kinds of disasters happened around the world. The modules that help in disaster
recovery and response are also surveyed and the techniques used are also deployed. From the deployed
issues and the techniques, comparative table of the works being surveyed are tabulated and the future works
are also suggested.
Keywords - Disaster Management; Social Data; Data Mining.
100
ISBN: 979-8-3503-4821-7
PAPER : CS743
Product Recommendation System using Deep Learning based Recurrent Neural Network
S.Shanthi , K. Nirmaladevi , Jayesh Krishna K ,Jayachandran P , Lalith Mohan S , R.Kuppuchamy
, S.Priyanka
[email protected]
ABSTRACT - Today's marketing strategies place a high priority on comprehending customer sentiments.
It will not only give businesses a better understanding of how their clients view their goods and/or services,
but it will also give them suggestions on how to enhance their offerings. We make an effort to comprehend
the relationship between various factors in customer reviews on an online store for women's clothing. The
proposed system also classifies each review, according to, whether it recommends the product in question
or not, and whether it has a positive, negative, or neutral sentiment. With the exception of review titles and
review texts, we used univariate and multivariate analyses on datasets features to accomplish these goals.
Additionally, we have implemented a bidirectional recurrent neural network (RNN) with a long-short term
memory unit (LSTM) for sentiment and recommendation.
Keywords - bidirectional recurrent neural network (RNN) with a long-short term memory unit (LSTM)
PAPER : CS747
Prediction of Crops Based on a Machine Learning Algorithm
Richa Kumari Karn , A. Suresh
[email protected]
ABSTRACT - Agriculture is crucial to any country's economy. Farmers around the world face a constant
challenge in trying to keep up with the rising demand for food crops in the face of fluctuating climates and
an alarming rise in population. One of the most widely grown cereals, wheat supplies a significant portion
of the world's main food supply. This heat-sensitive crop is being severely harmed by the unusual rise in
environmental temperature and decrease in the amount of rainfall. Scientists from all around the world have
been looking at what are called Climate Sustainable Agriculture Practices in an effort to boost wheat crop
yields while reducing the impact on the environment. Predicting crop yields before harvest can assist
scientists and farmers evaluate risks and implement preventative actions to maintain a consistent
agricultural harvest. There are two main types of models used to predict harvest yields: crop growth models
and data-driven models. The time, money, and accuracy costs associated with using crop growth models to
predict crop yields stem from the fact that these methods are sensitive to environmental variables. So, the
farmer can't do anything in the nick of time to boost his crop's production. With the advent of machine
learning algorithms, data-driven models have become even more effective at a fraction of the cost of
traditional empirical models. Machine learning and Machine learning have come a long way, but they
haven't been completely used for precise crop output forecasting. With this study, the authors want to
provide a reliable method for estimating future wheat harvests in one of India's Punjab provinces. For
precise and timely wheat crop yield prediction, an KNN with DT hybrid Machine learning model is
proposed. To further improve the model's performance, the researchers used a genetic algorithm to tune the
KNN-two DT's most crucial hyper parameters: the size of its window and the number of neurons in its
hidden layer. This study has also examined the impact of environmental factors on agricultural yield in
order to isolate the most important environmental parameters for future regulation and monitoring in order
to more accurately predict crop yield. The proposed model for predicting wheat crop yield was tested with
a battery of trials. The effectiveness of the suggested model was validated through a comparative
comparison with state-of-the-art approaches for yield prediction. Farmers, policymakers, and planners can
101
ISBN: 979-8-3503-4821-7
all benefit greatly from the proposed study by improving their ability to make informed decisions and take
corrective and preventative action to boost agricultural yields.
Keywords - Crop, Prediction, Machine learning, KNN, DT (Decision Tree).
PAPER : CS750
Movie Recommendation System
SHIVAM KUMAR GUPTA , Dr. A. SURESH
[email protected]
ABSTRACT - Because of the advent of the Internet, now have access to an abundance of data across
many disciplines. However, consumers frequently face situations where they have a plethora of options to
consider and could use some guidance navigating those options. An effective method for closing this gap
is the use of recommendation systems. There are many different methods being used to develop
recommender systems, but they can be broken down into two broad categories: content-based and
collaborative filtering. In order to address the limitations of traditional recommender systems, researchers
are increasingly turning to hybrid approaches, wherein multiple recommendation methods are combined.
In addition to these three tried-and-true methods, recommendation quality can also be enhanced by
employing a context-based recommender system. Several methods exist for modelling contextual
information in a recommendation system, including pre- and post-filtering, as well as contextual
modelling.In this paper propose a hybrid system that combines the best features of both pre- and post-
filtering contextualization techniques. In order to make better movie recommendations, the proposed
method will make use of a database that is rich in context and contains information such as user data, item
data, ratings, and contextual information. The proposed method as a whole is broken down into stages. To
generate the first set of recommendations, it will be first applying a contextual pre-filtering approach to the
entire database based on the most important contextual attribute for a user, thereby reducing the multi-
dimensional data into a reduced dataset. The recommendations are then sent to a contextual post filter,
where they undergo additional processing in the form of filtering and adjustment in light of the other two
pertinent contextual attributes for that user.
Keywords - Movie recommendation, Python, Analysis, Demographic filtering.
PAPER : CS751
ANALYSIS OF PERFORMANCE, SCALABILITY, AVAILABILITY AND SECURITY IN
DIFFERENT CLOUD ENVIRONMENTS FOR CLOUD COMPUTING
SANDHIYA V , Dr A. SURESH
[email protected]
ABSTRACT - Success and acclaim of cloud computing have been boosted by recent developments in the
field and in technology more generally. Cloud computing allows for flexible, on- demand service delivery.
Instead of using a server on a local area network (LAN) or a single computer, cloud computing takes
advantage of distributed, centralized, and highly sophisticated remote network sites and servers connected
to the internet for the purposes of storing, managing, and processing data. Statistics in cloud computing are
now being carefully preserved with an emphasis on studying data in the cloud and its many connected
security problems. Concerns about data loss in the cloud and the measures taken by different service
providers to address those concerns are also addressed. Cloud storage makes data easily accessible for a
wide variety of uses, but as the cloud grows in popularity and data distribution options expand to
accommodate a diverse user base, the privacy of that data becomes a more pressing concern. It necessitates
the covert transfer of data and the careful distribution of decryption keys between authorized parties. That's
102
ISBN: 979-8-3503-4821-7
why there are so many different approaches, and some of them centre on combining attribute-based
cryptography (ABC) and AES. While it's true that any given technology may have certain benefits, it's also
important to consider any drawbacks it may present, which may include things like increased demands on
the system's processing power and memory, as well as any potential bottlenecks in the network. Load
balancing refers to the process of distributing the workload over multiple nodes in a distributed system in
order to improve various aspects of the technology, such as the most efficient use of resources and the
speed with which jobs may be completed. Load balancing is used to ensure that all of the mainframes in
the building or all of the nodes in the network are doing nearly the same amount of work at any given
moment, hence preventing the aforementioned issues. Achieving this goal is crucial to developing a
practical and aesthetically pleasing loading algorithm. Now that we know the advantages and disadvantages
of various load balancing algorithms. The idea of FCS is used to construct a newly developed load
balancing algorithm.
Keywords - Cloud Storage, security, access control,cryptography,public key, data protection, privacy, risks
and threats, load balancing, throughput, latency.
PAPER : CS759
Sentiment Analysis Of Books Review Using Machine Learning
Dr.A.Tamilarasi , Mrs.K.Chitra , Dinesh Kumar.S , Karthick Raj.T, Kalaiarasi.V
ABSTRACT - Sentiment analysis plays a significant role in assessing people's emotions and feelings with
the use of data processing. Several examiners have offered strategies to precisely discern feelings in the
past. Sentiment analysis is done by using machine learning models for classification, text mining, text
analysis, data analysis, and data visualization to identify positive and negative commands in a natural
language processing challenge. The plain text of the commands will be cleaned and prepared beforehand.
We'll next extract numerical features from the data, which will be combined with the sentiments of the
commands to train models and detect sentiment. The purpose of this exercise is to find the best algorithm
for sentiment Analysis. We consider using a dataset of ratings and reviews on books that are given by
customers. The rating that customers give to books is used to analyze customers’ Sentiments. As a result,
the job is to give an algorithm that gives the best accuracy among all others. a) To predict the sentiment of
customers that which books they would like to prefer, based on reviews given to books. b) To carry out
"sentiment scrutiny" so that the review is judged as either positive or negative by considering the rating. c)
Determine the proportion of positive and negative reviews. d) Applying the Algorithms to the collected
Dataset.
Keywords - Naïve Bayes, Logistic regression, Classification report, Decision Tree, SVM .
PAPER : CS760
Anaemia Estimation for Patients Using a Machine Learning Model
Ms.G.Keerthiga , G.Sudheer , G Venkatesh , G Hemanth
[email protected]
ABSTRACT - Machine ailment hospitalization plans save time, improve precision, and completely
eliminate the need for additional employees in medical decision making. According to a few food studies,
one quarter of the state of the experience is weak. As a result, establishing a machine learning back
propagation algorithm fit important process includes installation is in high demand. The objective is to
figure out which classifier, or combination of classifiers, achieves the highest level of faultlessness in blood
cell placement for choruses’ discovery. To discover and estimate emptiness, Lasso and Ridge regressions
103
ISBN: 979-8-3503-4821-7
were used. In regards to accuracy, however, the classifier Ridge outperforms the Lasso reversion. As a
result, another more powerful invention should be obligated to indemnify for maximum accuracy.
Keywords - Detection, Machine learning, Back propagation algorithm, Lasso regression, Ridge regression,
Lasso reversion
PAPER : CS761
Unifrol - Versatile Online Campus Networking Platform for Students using Social Media
Architecture
Shivendra Pratap Singh , Dasu Siddharth , DR. S. Prabakeran
[email protected]
ABSTRACT - Owing to our ever-changing closeness with technology, particularly accelerated after the
pandemic period, students have shifted to having an almost inevitable form of presence in various online
communities, forums, social sites et cetera. Unifrol aims to constructively and organically redirect most of
this presence by spinning up a centralized & versatile semi-regulated platform for students, by students. It
is offered as a repurposable service to any university/college as a pre-configured space to effectively allow
students to also thrive online within their own college community. For this purpose, it will be designed
with facets to encourage additional horizons of growth and interactions for students, promoting more intra-
community participation in clubs, university bodies, micro-economies and markets as well as fostering
better awareness through calibrated communication channels. Additionally, it will also serve to break down
non-essential & cumbersome barriers among student and management bodies. Mostly anonymized
platform with a catalyzing mix of an unofficial/free online presence would promote more natural platform
engagement among students, and in-turn, much more efficient interactions among students, university
management, and even some external stakeholders
Keywords - Anonymized, Social, Engagement, Content, Interactions, Campus
PAPER : CS762
CHATBOT FOR HEALTHCARE USING MACHINE LEARNING
Kaladevi R, K.M Nithiya Sri , S Saidineesha , S Sai Gayatri , P Keerthi Priya

[email protected]
ABSTRACT - Healthcare has become an important part of living a healthy life in today’s society, but
consulting with a doctor for every health problem is very difficult for common people. The Medical chat
bot is created to diagnose diseases and provide basic information before consulting a physician. By doing
so, healthcare costs will be reduced and medical knowledge will improve. A chat bot interacts with users
through natural language. The chat bot then saves the information in a database, chooses sentence
keywords, decides whether to answer an inquiry-, and provides information. With this approach, people
can spend less time in hospitals while performing repetitive tasks like providing solutions, sending emails,
marketing, and analyzing results. Similarly, calculations are done using n- grams, TF-IDFs, and cosine. So,
this gives the best result for detecting the disease through a medical chat bot.
Keywords - Chat bot, Machine Learning, Natural LanguageProcessing, Health Care, TFID, N-gram,
Virtual Assistance.
104
ISBN: 979-8-3503-4821-7
PAPER : CS765
Machine Learning based Intelligent Career Counselling Chatbot (ICCC)
Dr. Reema Goyal , Navneet Chaudhry , Mandeep Singh
[email protected]
ABSTRACT - The Intelligent Career Counselling Chatbot (ICCC) is proposed in this paper that provides
career guidance to 10th and 12th standard students for clearing their doubts and for learning purpose. It is
also helpful for choosing a career for higher studies. A special module has been developed for CSE and IT
domain of BTech students to provide help in their studies and guidance for choosing career for their higher
studies. The user is prompted by the chatbot with a variety of questions, to which the user must respond
with 1 for true and 0 for false. Through these questions, the chatbot defines a potential future event or result
based on the user's input and displays it on the GUI. Apart from career counselling, it also communicates
with users. The result developed for this chatbot consists of four major branches that are training chatbot,
adding functionalities like adaptors and other modules, machine learning algorithms for predictions, and a
server for GUI-based chatbot. The chatbot can be used for counselling different students to help them
finding their interests and to choose their careers accordingly.
Keywords - career counselling chatbot, chatbot, career guidance, chatterbot, conversational agents, machine
learning agent.
PAPER : CS766
A Review on Long Short Term Memory based Stock Price Prediction Techniques
Kandlagunta Divyavalli , P Raghuraman
[email protected]
ABSTRACT - The stock market is a widely-known financial vehicle with a significant influence on the
global economy. It will be every investor's ambition to be able to foresee the market's rise and fall and
profit handsomely from doing so. However, the future of prices is quite uncertain. Even though several
types of analysis such as fundamental analysis, technical analysis, etc. have existed for quite some time.
Today, several algorithms are utilized to make predictions about prices in the future. The ability to
accurately forecast future stock prices is crucial since it provides traders and investors with valuable
information about the market's potential direction. Therefore, they will be able to enter or leave the market
at the optimal moment. The most difficult endeavor is to foresee the stock market. There are many aspects
to consider, both objective and subjective, rational and illogical. The combination of these elements makes
it not only difficult but impossible, to make an accurate prediction of the price of shares. As part of our
research, we analyzed several Neural Network techniques, including Long Short-Term Memory (LSTM)
and Recurrent Neural Networks (RNN), both of which are effective in predicting stock price.
Keywords - Stocks, Machine Learning, Variability, Price Fluctuations, Trends, Patterns.
PAPER : CS769
Big Data – Enhancement in Sports : Exploring the Impact of Technology Integration
Dr. S. Venugopal
[email protected]
ABSTRACT - There are fewer technologies available that aid in player development, the usage of
technologies to enhance athletic performance has expanded. The adoption of technology may aid in players'
expansion. The current technology has been used in various fields, including health and wellness,
psychosocial concerns, and transitional skills. The paper aids in study of technology influence players'
105
ISBN: 979-8-3503-4821-7
capacity for athletic competition. Sports is the best methods to gauge someone's athletic prowess, but
technology can still be useful. Where technology can flourish people may fall short in administering and
officiating sports, technology can succeed. It is possible to determine a player's effectiveness and
consistency over the course of the season by using this information. This can lead to predictive modelling,
which gives analysts the ability to judge if a player is a one-season wonder or whether their performance
is becoming better over time. Strategizing is a crucial part of any activity, team or individual sport included.
Professional athletes and teams must employ these strategies in order to compete and out perform their
rivals. Modern coaches use big data sets to create efficient procedures that will help both individual athletes
and the team as a whole. Data science gives coaches the ability to create highly individualised athlete
matches and other game plans for each game the team plays, especially for professional teams. In this way,
the team's tactics will continue to be unpredictable effective. Bigdata is used effectively in sports and this
technology integration is used to achieve more in this field.
Keywords - Sports, Technology, Bigdata, Skills, Bigdata, Tools .
PAPER : CS774
Image Captioning Using Deep Learning Techniques for Partially Impaired People
Suganthe Ravi Chandaran, Shanthi Natesan, Geetha Muthusamy, Pravin Kumar Sivakumar, Prema
Mohanraj, Richard Joyal Gnanaprakasam
[email protected]
ABSTRACT - Image captioning is used to describe an image based on the features and actions that are
present in that image. The existing image captioning primarily uses an encoding and decoding structure,
with the encoder which extracts image features using CNN model as well as the decoder uses LSTM model.
The current encoding and decoding scheme makes extensive use of the attention mechanism. Gradient
explosion is a problem with the current image caption models, which are based on recurrent and
convolutional neural networks, and therefore are not particularly good at extracting valuable information
from images. To overcome this problem, the YOLOv5 and Bidirectional LSTM model is proposed.
YOLOv5 is used to identify the objects in a given image and a bidirectional LSTM (Bi-LSTM) layer is
used to extract the features of the given image. This proposed algorithm gives the optimized result with
good accuracy. The Flickr8k benchmark dataset is used to test this approach. From the results it is known
that the trained model performs better than alternative encoder-decoder methods that depend solely on
global image features. The metric which has been used to evaluate the model is BLEU which is mainly
used for machine translated text evaluation. This model gave a 0.7 BLEU score.
Keywords - component; Image Captioning, Deep Learning, Bi-LSTM with YOLOv5, Caption Generator
.
PAPER : CS775
Artificial Intelligence Application for Effective Customer Relationship Management
Dr. Mohammad Salameh Almahairah
[email protected]
ABSTRACT - Machine learning and artificial intelligence (AI) allow businesses to make better use of the
data they already have and to find solutions to difficult problems, both of which make it possible for them
to expand their operations in a timely and cost-effective manner. In recent years, artificial intelligence (AI)
and process automation have emerged as two of the most sought-after innovations in customer relationship
management (CRM). This is because businesses put in significantly more effort to guarantee complete and
total happiness for their customers (CRM). AI in CRM is aimed to decrease the amount of boring, routine
106
ISBN: 979-8-3503-4821-7
labour that people must do by analysing standard operations and freeing up time for employees to focus on
things that are more important and demanding. With the use of artificial intelligence (AI), customer
relationship management software may streamline and automate a variety of labour-intensive
responsibilities, including following up with clients, sending thank-you messages, scheduling
appointments, and managing calendars. When artificial intelligence (AI) is linked with typical customer
relationship management (CRM) systems, it is anticipated that AI would increase lead creation and
visibility and improve customer experience, procedures, and team productivity. In customer relationship
management, the goal of artificial intelligence is to do more than gather data; this has the potential to
increase sales and keep existing customers satisfied.
Keywords - Artificial Intelligence, Machine Learning Techniques, Customer Relationship, Management.
PAPER :CS786
AUDIO SUMMARIZATION OF PODCASTS IN REAL-TIME
R. Aishwarya, R. Yogitha, Gorli sathvik sri harsha, Barma Bharath.
[email protected]
ABSTRACT--- The COVID-19 pandemic has compelled most government work to be done online. There
are business meetings, primary health consultations, job interviews, and online colleges all across the
world. Google Meet, Microsoft Team, and other online meeting tools are frequently available. This study
can be put to a lot of good use. This document describes how to extract information from a recorded video
and provide a written or speech report. The technique can be used to create meeting minutes, stories,
subtitles, and instructional notes using entertainment videos. The proposed method converts audio from
video to text. Additionally, text summarising algorithms were applied. The text summary can be used
separately or in conjunction with audio by system users. YouTube videos are used for testing, and Python
is used to implement the strategy. The suggested method is manually evaluated on the provided video
collection because there isn't an appropriate dataset or efficiency benchmark.
Keywords: Natural language processing , Machine learning, python speech recognition, TF-IDF,
ASR(Automatic speech recognition), Summarization.
PAPER CS789
Smart Autonomous Voice Control Obstacle Avoiding Robot
Ashok Reddy Kandula, Poojitha Pagolu, Kalyanapu Srinivas, Sindhu Pasupuleti, Sri Rama Krishna
Datta Movva, Sai Charan Nancharla
[email protected]
Abstract—The study aims to build voice-controlled robots with autonomous braking, obstacle avoidance,
and speed reduction capabilities. This research will enable the development of robots that can accurately
and autonomously assess and respond to environmental conditions, such as obstacles in their path. The
primary objective of the robot is to use voice commands to steer the robotic vehicle and analyze speech
input to carry out the necessary actions to avoid obstacles. The research is expected to significantly improve
the safety of autonomous vehicles by allowing them to detect and respond in real time to changes in their
107
ISBN: 979-8-3503-4821-7
environment. The robot is voice-activated and connected to an Android smartphone through Bluetooth.
The robot will also be able to recognize obstacles with the help of an ultrasonic sensor. The physically
disabled and the primarily blind will benefit from further developments in this technology, among many
other groups of people. This technology can fundamentally alter how people view mobility and provide
many opportunities to those who would otherwise need to rely on other modes of transportation.
Keywords- Obstacle Avoiding, Voice Control, Speech Recognition, Autonomous System, Ultrasonic
Sensor, Bluetooth Control System.
PAPER :CS799
AN ENHANCED YOLO BASED TRAFFIC ANALYSIS SYSTEM
Marella Naga Chaitanya Sneha,Thota Satya Veera Manikanta,Battula Greeshma,Praveen
Tumuluru,Valini Sunthwal,K. B. V. Brahma Rao.
[email protected]
Abstract—As the number of automobiles on the road rises daily, population increases so traffic is the main
problem faced by the people in our society. Traffic analysis is a task that help the user to avoid route that
are going to pack with traffic jam in cities. Most of the cities suffer from traffic jams so implementing a
tools that helps to predict the traffic jam and take less time to reach their destination To control this traffic
problem many are implementing various machine learning algorithms which are very effective in
controlling the flow of traffic in various places some algorithms like SVM (Support Vector Machine), and
other CNN (Convolutional Neural Networks), LSTM, XGBOOST, YOLO (You Only Look Once), RCNN,
RNN, SSD Machine Learning Algorithms etc. It can increase traffic operation effectively, cut carbon
emission, ease traffic congestion, and aid road users in making smarter travel selection.
Keywords—YOLO, CNN, SVM, RCNN, RNN
PAPER:CS800
Distributed Inference Approach on Massive datasets using MapReduce
M.Priyadarshini, M.K. Dharani
[email protected]
Abstract—Contemporary computer systems and applications generate high volume of data every day.
Gaining knowledge from this ever-growing high velocity and high volume data is crucial to have insights
and business intelligence. Using semantic web approaches for generating inferences to gain knowledge
have been quite successful. When processing large amounts of data, a centralised method for finding
inferences in ontologies will be ineffective. Therefore, to solve this problem, a distributed strategy is
needed. The major challenges on large scale data are the difficulty in deriving suitable triples for
appropriate inferences, to reduce the time spent in processing of inference and the requirement of scalable
computation capabilities for large dataset. Also, storage space for increasing data must be addressed
efficiently. This paper proposes a distributed conjecture approach to address the above issues by
construction of SIM (Sparse Index Method) and ATC (Assertional Triples Construction) and to efficiently
process the users’ queries.
Keywords— Big data, Semantic Web, MapReduce, ontology reasoning, RDF.
108
ISBN: 979-8-3503-4821-7
PAPER:CS803
Continuous Digital System Analysis in different System Softwares Using Keyloggers to Validate the
need of Security
Srikakulapu Bhavitha,Matta Chenna Rao,Nancharaiah,Dr.S. Suhasini
[email protected]
Abstract— In the world with increasing technology, Security is the ultimate factor that has to be
maintained. In this regard, one single vulnerability is all an attacker needs to cause damage. However, it is
sometimes necessary to make use of such vulnerabilities in order to monitor the usage of
desktop/laptop/mobile devices for information security. The personal work of various customers is now
being monitored by a number of businesses through different web applications. The motive of this model
is to study the working of a keylogger by building a real-time working keylogger as well as a clipboard
logger. that captures user entered keystrokes and copied text data respectively. This recent pandemic has
affected the entire society where most of the people started using electronic devices for various purposes.
The purposes can be good or bad based on the users’ intentions ranging from student’s usage for education,
anti-nationals using devices as communication channels with their networks to create catastrophic events,
children watching adult content, employees stealing workplace information etc. As a result, the high usage
of the systems requires the necessity to keep a close check on the users in order to avoid these illegal
activities from happening. Accordingly, this monitoring can be useful for employers to observe their
employee system usage and can help the parents to examine their children’s system usage. Thus, a
keylogger can be used to capture the keystrokes and store the user data so that the higher authority can
supervise this usage and act appropriately to circumvent any forbidden actions. However, there is a chance
that the users can cleverly manipulate this keylogger mechanism by copying text in the device and
performing actions on that info. So, the model is extended to use a clipboard logger that collects the
clipboard data to avoid any discrepancies caused by the keylogger data. This results in the overall
surveillance of the system.
Keywords—Clipboard logger, Information Security, Keylogger, Keystrokes
PAPER:CS813
Fever Detection System Using Machine Learning Algorithm
Subramanian K,Navneeth krishnan B ,Nushanth S,Kaladevi R
[email protected]
Abstract—Worldwide many new diseases are spreading and causing several illnesses to the people. Most
of the diseases are difficult to predict and lead to life threat. Fever is one among the diseases. Even though
it’s a familiar disease, nowadays different kinds of fevers are spreading across the environment. Either
Fever can be direct transmission or due to any side effects. Detecting the fever on time helps to prevent
disease. Diagnosis of undifferentiated fever is a difficult task for the doctors. Availability of data is a big
issue. Obtained data also not in the required format. Machine learning is one of the tools to predict the
fever. In this proposed system, real time data is collected and a logistic regression algorithm is used with
the accuracy of 98. Logistic regression is a statistical method used to predict the outcome of a dependent
variable based on previous observations.
Keywords: Fever Prediction, Machine learning, Logistic Regression, real time data.
109
ISBN: 979-8-3503-4821-7
PAPER:CS815
Maximising Efficiency of Network Resources in SDN through Optimal Packet Routing using
Reinforcement Learning
Anush K Byakodi, Mohammad Zakee Hungund, Aman Khan N Athani, Khadar Wali B Afghan,
Somashekar Patil, Narayan D. G.
[email protected]
Abstract—Software-defined networking (SDN) has increased in popularity in recent years owing to its
ability to centralize network device control. This is the cornerstone of data center networks, and as more
IoT-enabled devices are added, it is expected to skyrocket. The employment of Machine Intelligence to
handle and route the majority of traffic can enhance centralized SDN-based control, allowing for high and
occasionally subpar network management. Owing to the development of Reinforcement Learning (RL)
combined with Deep Learning, which allows autonomous flow management in SDN, intelligent routing
has gained popularity. However, the existing SDN routing algorithm has poor link usage and is unable to
update and change according to real-time network conditions, making it challenging to forecast and predict
how communication networks will evolve in the future. Consequently, we present a routing optimization
strategy based on Quality of Service (QoS) variables for SDN-based Data Center Networks using Deep
Reinforcement Learning. We provide a reinforcement learning-based routing method for SDN that
outperforms Dijkstra’s shortest path in terms of the throughput and average latency. In the proposed
method, we integrated SDN with RL, which takes input from the state in which the current network is and
routes traffic to achieve the maximum QoS parameters. The parameters considered were the throughput
and delay. The presented solution is a network- oriented, self-learning routing system named DeepQoSR,
based on Advantage Actor-Critic Networks. We compared our proposed technique with traditional
algorithms, such as the shortest path, and present our analysis.
Keywords—Software Defined Networks, Data Center Net- works, Deep Reinforcement Learning, Machine
Learning, Quality of Service (QoS), Advantage Actor-Critic.
PAPER:CS817
Secure the smart grid by machine learning
Srinivas Kolli ,Hasan M ,Bramah Hazela ,A.Azhagu Jaisudhan Pazhani
[email protected]
Abstract— Smart Grid Technology has to tighten its security because of the growing threat of
cyberattacks. Machine learning approaches are more effective in detecting assaults than conventional
methods. The difficulties of detecting hostile activity and infiltration in intelligent grid communication
networks were investigated with machine learning techniques. In this work, we apply machine learning
methods to determine if the measurements came from an attack or a secure environment. By using several
machine learning methods, including Perceptron, Logistic Regression, Support Vector Machine, and the
K-Nearest Neighbors algorithm, the suggested system hope to foresee the occurrence of erroneous data
injections in the smart grid.
Keywords- smart grid, cyberattacks, machine learning techniques
110
ISBN: 979-8-3503-4821-7
PAPER:CS818
Assessment Based On IoT For Efficient Information Surveillance Regarding Harmful Strikes Upon
Financial Collection
Jehan Kadhim shareef Al-safi,Dr Atul Bansal ,Dr Mohd Aarif ,Dr.Mohammad Salameh Zaid
Almahairah ,Geetha Manoharan,Firas Jamil Alotoum.
[email protected]
Abstract — Several more individuals consider the internet of things to be the upcoming big idea, however,
they are woefully ignorant of its standards as well as advantages. It is believed that by the finish of 2022,
anywhere from 50 billion to 60 billion internets things will have been implemented in our rapidly expanding
globe. A large number of advantages of the internet of things in history's rapidly expanding and operating
globe. As everyone understands, the internet of things has presently evolved into an essential component
of our creation, and it's experiencing quick advancements in the field. Now that our subject has informed
us that innumerable threats have been carried out against the internet of things, we will talk about how the
internet of things must be handled or supervised so that an efficient tracking of the information can be
carried out, as well as how it must create a secure space or evaluation of the information, and how it ought
to defend itself against suspicious threats carried out again on financial collectibles as well as companies.
Keywords — IoT, Data processing, Data Management, Device Architecture, Gateway, Wireless Network,
Performance, Security
PAPER:CS819
An In-Depth Analysis of the Role That ML and Big Data Play in Driving Digital Marketing's
Paradigm Shift
Mano Ashish Tripathi ,Dr Femmy Effendy ,Dr. John Paul.M ,Ravindra Tripathi ,Geetha
Manoharan ,r. Mohd Aarif
[email protected]
Abstract— Machine learning (ML) is an artificial neural network (ANN) that helps developers improve
their software's predictive abilities before they have all the data they need. Because information is so
priceless, progress toward fully autonomous agents requires better methods for managing the omnipresent
content infrastructures that exist today. All sorts of fields have benefited from advancements in computer
vision and AI, from medical diagnosis to data presentation and operations to scientific study, and so on.
Learning from polluted or erroneous data may be expensive, much as training for a sport can be dangerous
to those who are vulnerable to injury. An organisation will incur costs rather than see benefits if its
algorithms are improperly taught, as explained in Approaching Data Science. Organisations need to be able
to verify the quality and consistency of any large datasets, as well as their sources, to ensure the efficacy
of any algorithm.
Keywords- machine learning, Data Acquisition, Data processing, Algorithm, Data Management,
Interpretation, statistics, probability, supervised learning, classification, regression, clustering, big data,
digital transformation, inquiry, Data protection., Security
111
ISBN: 979-8-3503-4821-7
PAPER:CS821
An Elaborate Review of Diagnosing Glaucoma Disease Using Machine Learning and Deep Learning
Techniques
R.Thamilselvan, P.Natesan, R.R.Rajalaxm,S.Mohan, S.Muruganantham , R.Vanitha.
[email protected]
Abstract— The vision of a human is very important in their life. In today’s world lot of medical
advancements have arrived to treat all the diseases. In this part the earlier identification of the disease is
still a challenging part. In this review the study path mainly focus on the eye disease. The major vision
problem with eye sight is due to cataract and glaucoma. It is curable if we take utmost care in the early
stage of caused by constant checkup. Glaucoma is an optical problem caused by the neuropathy due to loss
of retinal ganglion cells. This Glaucoma can be detected through fundus images, which will help to identify
the vision loss at the earlier stage. Optic Nerve Head (ONH) was the first parts that get damaged by the
glaucoma so its diagnosis will always focus on the parts such as Optic cup and Optic Disc in the retina. In
this review work the datasets based on these were collected and study was made based on the various
machine learning and deep learning techniques.
Keywords- Fundus Image, Optic Nerve Head, Optic cup, Cup- to-disc Ratio, Optic Disc, Glaucoma.
PAPER:CS822
Analyses of Face Mask Detection Using Deep Neural

Siddarthha Yelchuri,R.Sumathi,Krishna Murthi Vavilala ,Krishna Sai Mahesh Vellampalli ,
K.Yashas
[email protected]
Abstract- Corona virus started and spreading throughout the world and many peoples are affected from
2019 to 2021. To prevent the spreading government ask all the citizens to wear the mask without fail in all
public places. However, some people are not wearing masks as the spread of coronavirus increasing day
by day. To avoid these problems, it is necessary to look for people who are not wearing masks in public
places, but there is a problem that those who check may be infected with coronavirus. In order to solve this
problem, we need to implement an automatic facemask detector by using deep neural network algorithm
for monitor the proper wearing of mask in all places and detect the improper wearing in accurate manner.
We yield 97% detection accuracy whereas KNN yields 95% and BPNN yields 96%.
Keywords- CoronaVirus, face Mask detector, deep learning, deep neural network.
PAPER:CS825
MOVIE RECOMMENDATION SYSTEM USING CONTENT BASED NOVEL FEATURE
EXTRACTION INFORMATION
Nagaraj P,Sumanth Kumar Reddy P,Venkata Rao K,Mohan R,Jaswanth V,Chethan Kumar M,K.
Muthamil Sudar.
[email protected]
Abstract- Content-Based Filtering and Collaborative Filtering are the two key techniques used by
recommender systems to help them provide relevant recommendations. In order to create a report for this
paper, we developed a hybrid technique that benefits from both content and collaborative filtering
112
ISBN: 979-8-3503-4821-7
algorithms. The algorithm discussed in this article differs from other work on the topic since it uses a
cutting-edge technique to identify content similarities between two items. The study in the paper shows
why this new technique is valid and how it can result in useful suggestions. On existing person and object
data, the aforementioned strategy outperformed the two most popular approaches, Pure Collaborative
Filtering and Singular Value Decomposition, in terms of outcomes.
Keywords:-Movie Recommender System, Collaborative Filtering, Content Based Filtering, Hybrid

Approach, Singular Value Decomposition (SVD).
PAPER:CS827
Youtube Comment Spam Detection
P Nagaraj,P Girish Kumar Reddy,K. Muthamil Sudar,SK Baji Babu,P Thrived,P Siva Rama
Krishna.
[email protected]
Abstract— YouTube is the most commonly used application these days. There are both good users and
bad users who use youtube. Spammers have realized that these types of social media are easier to lure the
users into malicious activities. In the current project, spam detection was performed on selected youtube
comments and we predict the entered comment as ‘SPAM’ or ‘NOT SPAM’. Google Safe Search and
Youtube Bookmaker are two tools that can be used to block harmful links; however, they cannot
immediately safeguard real-time users. Four Artificial Intelligence approaches have been employed by
victimization to run the survey for the spam comment detection methodology: Call tree, Ada Boost, logistic
regression, and Random Forest. We shall achieve a 91.65 percent associate degree of accuracy by using
neural networks and also to beat the current action course by around 18%. The difficulty of spam and the
efficiency of the most well-known AI approaches (SVMs, Bayesian representation, ANNs, k-NN).
Keywords—spam, youtube, social media, comments, predict
PAPER:CS830
Predicting User Click Behavior on Social Media Ads Using Machine Learning
Ashish Kumar, Divyanshu kishan, Harshit Kandpal, Himanshu Saraswat, Jyotiraditya Singh
[email protected]
Abstract—Social media advertising has become an increasingly important aspect of modern marketing
strategies, with platforms like Facebook, Instagram, and Twitter offering targeted advertising options to
businesses. One key metric for assessing the effectiveness of a social media ad campaign is the number of
clicks it receives, but predicting whether or not a user will click on an ad is even more valuable for
businesses. In this study, we aim to explore the use of machine learning techniques for predicting user ad
click behaviour on social media platforms. To do this, we collected a dataset of real-world social media ads
and applied various machine-learning algorithms to build predictive models. Our results show that certain
algorithms, such as logistic regression and support vector machines, perform well in predicting whether or
not a user will click on an ad. We also identified important factors that influence ad click rates, such as the
ad copy and the target audience. In addition to the predictive models, we conducted a thorough analysis of
the dataset to understand the underlying patterns and trends that influence ad click behaviour. This included
an examination of the relationship between different features, such as age, gender, and the likelihood of a
user clicking on an ad. Overall, our study provides valuable insights for businesses looking to optimise
113
ISBN: 979-8-3503-4821-7
their social media ad campaigns and for researchers interested in understanding the factors that influence
ad performance. By understanding the factors that drive user ad click behaviour, businesses can better target
their advertising efforts and improve the effectiveness of their campaigns.
Keywords -social media advertisement, social networking, machine learning, classification, data analysis
PAPER:CS834
A Technique for Digital Image Watermarking That Detects and Restores Tampering
Pavan. A. C, Dr. M.T. Somashekara
[email protected]
Abstract - In addition to ensuring image authenticity, watermarking aids in the recovery of digital images
that have been altered or tampered with during transmission. Although many techniques have been
developed for embedding the watermark that results in recovering itself of photos, they only effectively
work in the uncompressed multiple domains, leaving images vulnerable to attacks on image compression,
such as noise addition. In order to effectively detect tampered areas and assure self-recovery of JPEG
images, we offer a solution to address these issues in the JPEG domain. Source code compresses the original
image, while channel code protects it from tampering. As a result, the image will be compressed and then
used as the watermark which is embedded. The approach proposed is successfully combined with these
JPEG images, producing increased image quality as well as strong performance against noise attacks.
Keywords - Digital image watermarking, LDPC, Detection of tampering, SPIHT, JPEG Image
Compression
PAPER:CS836
Automatic Extractive Text Summarization of Hindi Text using Deep learning approach
Nandini Kapoor,Ayush Gupta,K.Meenakshi .
[email protected]
Abstract— In order to swiftly sift through the vast amounts of textual material available online, Automatic
Text Summarization (ATS) is in great demand. In this research, the Real Coded Genetic Algorithm (RCGA)
is used to the Hindi movie reviews available on the Kaggle dataset in order to suggest an ATS approach
for the Hindi language. The approach consists of five stages: pre-processing, extraction of features,
processing, paragraph ranking, and summary output. In a rigorous research on many feature sets, sentence
similarities and semantic segmentation characteristics are merged with some other features to produce the
evaluation metrics. Different compression rates are evaluated in order to extract the sentences with the
greatest scores as the corpus summaries. The ATS extractive approach provides a summary reduction of
65% when compared to current summarization methods. A text summary tool condenses the text and shows
the user only the crucial information. The significant sentences are chosen via the extraction approach
based on a theme approach. Hindi stop-words were eliminated before choosing thematic terms, and the
stemming procedure was used to find the sentences' root words. Stop-word removal removes useless words
from the input material, while stemming helps group together words that have the same numerical term.
The method relies on how frequently the radix of theme terms appear in the sentences to determine a score
for them. Sentences having the highest ratings are prioritized for inclusion in the summary. In order to
make the produced summary more similar to human-generated summary, it is then further treated based on
the elimination of superfluous terms from the selected summary sentences.
Keywords —Automatic Text Summarization, Genetic Algorithm, real coded genetic algorithm, TF-IDF
114
ISBN: 979-8-3503-4821-7
PAPER:CS838
Automatic Correction of Text Using Probabilistic Error Approach
Nagaraj P,Muneeswaran V ,Niamatullah Ghous,Mahtab Ahmad ,Pavan Kumar ,Vinayak
[email protected]
Abstract—Continuous research effort has been paid to automatic spelling correction. Although each study
provides a quick introduction to the problem, there is a paucity of work that would combine the analytical
foundations and provide a summary of the approaches investigated thus far. Our study has chosen spelling
correction publications from 1991 to 2019 that were indexed in Scopus and Web of Science. The first group
follows a predetermined set of guidelines. The second group makes use of an extra context model. The
third subset of the survey's computerised spelling checkers can modify its model to fit the situation. The
summary tables identify each system's application domain, language, context model, and string metrics. To
fix the publicly available construction data, this work attempts to create an automated data correction
system. The handling of the construction data is hampered by its unstructured character. Due to different
types of users, lengthy system operation, and substantial pretraining time, the information management
system has a lot of data in inconsistent formats or even erroneous data. The conversion of the construction
data into a machine-readable format requires a lot of time and work
PAPER:CS839
COVID-19 post-vaccination adverse effects prediction with supervised machine learning models
Vaishali Ravindranath ,Sarojini Balakrishnan
[email protected]
Abstract—This electronic Pharmacovigilance with AI helps to trace possible adverse events among the
vaccinated population. The symptom pattern discovery from the vaccinated population obtained through
post-vaccination surveys provides insights for medical practitioners to study the possibility of adverse
events among vulnerable populations. This work contains post-vaccination survey data from an Indonesian
national referral hospital with 840 instances, 6 categorical input features, and 15 binary target attributes.
As there were multiple symptoms as a target, multi-target classification algorithms experimented on the
dataset. The inadequate sample size resulted in poor performance of the algorithm. To improve model
prediction performance, the target was converted into binary format. The population that exhibited at least
one symptom is considered symptomatic by the binary classification models. The supervised machine
learning model of test train split (80% - 20%) produced 89% accuracy with a decision tree classification
algorithm in the classification of symptomatic or non-symptomatic patients.
Keywords- Machine learning; Pharmacovigilance; Supervised algorithm; multi-target classification
PAPER:CS840
AUTOMATIC EGG INCUBATOR USING IOT
Mrs.K.Chitra ,Mr. S.Hari vishnu ,Mrs.T.Kalpana,Mr. K.Kishore Dr.M.Hemalatha,Mr. S.Manoj
[email protected].
Abstract---An egg incubator is a machine that mimics incubation by keeping the eggs warm at a particular
temperature range and at the right humidity with the aid of a hatching mechanism. An enhanced egg
incubator system that achieves excellent hatchability and can hatch a huge number of chickens at once is
developed by a well. This incubator system, which uses an Arduino microcontroller, has automatic
temperature, humidity, and egg-reversing capabilities. The GSM system can also assist farmers in
115
ISBN: 979-8-3503-4821-7
monitoring the smart incubator from a distance. The eggs smart incubator can be used to hatch the eggs at
their own location and will produce better results than the traditional approach, which results in an egg
hatch rate of 90.55%, a defect hatch rate of 0.41%, a batch hatch rate of 84%, and an egg hatch rate of
10.20% per batch of 50 eggs. Egg hatching is the main priority in agricultural areas.
Keywords—Microcontroller, Arduino, DTH, Global System for Mobile communication) GSM module,
can be added to improve it.
PAPER:CS841
Covid-19 Detection Using CNN Model with CT Scan Images
Gadde Lohith Sai Akhilesh,Kollipara Lokesh,D. V. Pavan Manikanta Teja,Putta Jashwanth
Sai,Praveen Tumuluru, Dr.S.Hrushikesava Raju
[email protected]
Abstract -Corona virus, a serious respiratory viral disease, has been spreading fast all over the world since
the beginning of 2020. Over 30 million individuals and even more than normal rate have contracted the
illness, and millions have died as a result. Despite the fact that there are various vaccines available right
now, positive cases are still on the rise. The virus can spread in tiny liquid particles from the lips or nose
of an infected person when they speak, sing, sneeze, cough, or breathe. These particles are various, ranging
from larger respiratory droplets to small aerosols. We will create a CNN (Convolutional Neural Networks)
model to predict if a person has covid or not using CT (Computed Tomography) scan images because the
diagnostic kits and related resources are insufficient due to the increase in covid cases. A CT scan is a
helpful diagnostic tool for identifying illnesses and injuries. It creates a 3D image of the soft tissues and
bones using a succession of X-rays and a computer. In a hospital or imaging facility, you might receive a
CT scan. Using the developed models, covid-19 was identified. As a fast and significant method to identify
covid-19, doctors can use CT scan images in conjunction with automated identification as coronavirus.
Keywords— Covid, Convolutional Neural Networks, Respiratory, Aerosols, Computed tomography,
Diagnosis
PAPER:CS842
Artificial Special Visual Geometry Group-16(VGG) Learning Model for Analysing Accuracy and
Precision of SARS-COV-2 Forecasting
Valaboju Shiva Kumar Chary,Venkata Naga Sai Suraj Pasupuleti ,Bellamkonda Satya Sai
Venkateswarlu ,Vijaya Babu Burra ,Saketh Vemuri,Praveen Tumuluru
[email protected]
Abstract— Current research attempts to predict SARS-COV-2 efficiently applying recent artificial
intelligence algorithms and lung radiography. In this investigation, we applied the promising Visual
Geometry Group16 transfer learning technique to identify SARS-COV-2 more reliably and fast. The
technique splits the lung radiograph image into two groups: SARS-COV-2 and normal. The Accuracy,
Exactness, Recognize, Support, and Harmonic Mean metrics are used to evaluate the model's effectiveness.
A total of 2000 radiograph samples were used in the study. The recommended VGG16 model provides an
outstanding calibration for identification of 0.93% for the knowledge involved in the given random sample,
which is greater than any current technique reported even in prior studies. Because the proposed method is
highly successful and accurate, it might be used to help and educate radiologists and other medical
practitioners in diagnosing SARS-COV-2 using lung radiographs.
Keywords— SARS-COV-2, Visual Geometry Group16, transfer learning
116
ISBN: 979-8-3503-4821-7
PAPER:CS845
Analysis of Cloud Computing Security Challenges and Threats for Resolving Data Breach Issues
Manmeet Kaur,Athira B Kaimal
[email protected]
Abstract— Technology now affects every element of the high-tech society, which creates an environment
that permits attackers to easily and cheaply gain access to and attack a particular organisation. The
organisation struggles to quickly notice it, though. With cloud computing, users can save data online and
over a network. Additionally, the data is accessible from anywhere in the world. IT-related capabilities can
be accessed by multiple users as a service through the internet because of cloud computing, and these
services are charged as per use. A substantial rise in the use of various cloud services has occurred based
on the adoption of cloud technology by several cloud computing providers, including Microsoft, Google,
IBM, Yahoo & Amazon. Amazon is ruling the market due its architectural characteristics. Different paid
tools and open-source tools are being developed to meet the requirements. Since security is a crucial matter,
most individuals are apprehensive about storing their sensitive information in a cloud, especially those who
are unaware about the rise in attacks and the advancement of technology. There have been numerous
advancements made when it comes to cloud computing, many challenges still need to be resolved, including
those relating to security, interoperability, resource scheduling, virtualization, etc.
This paper carries out a systematic literature review by detecting cloud computing security challenges &
threats and offers ideas for resolving data breach issues.
Keywords— Cloud computing security, cloud computing implementation, cybersecurity, cloud computing
management
PAPER:CS846
Web Information Retrieval using Vector Space Model and Docu-Tally Metric
Nitish Chaturvedi,Meenakshi K,
[email protected]
Abstract—Regardless of the type of data set, it is frequently challenging to sift through the vast amount
of data that is available on the Internet as a result of technological improvements. To deal with challenges
mentioned above, we have come up with a ranking approach which is computed using NLP and vector
space model. The approaches used for information retrieval start with a basic machine learning model and
progress to multi- stage architectures and frameworks like language modelling and term matching. The
main goal of this work is to use a standard retrieval process to glean insights from large amounts of data,
which is the problem we are aiming to solve. The method utilised in the study is latent semantic analysis,
which takes advantage of the semantic aspects at play and can be used to glean insights from lengthy texts.
Keywords—information retrieval, language modelling, latent semantic analysis, vector space model
PAPER:CS849
BLOCKCHAIN ASSISTED CLOUD STORAGE FOR ELECTRONIC HEALTH RECORDS
SANDEEP KUMAR E V,Dr. A. SURESH
[email protected]
ABSTRACT–The Secure Health Framework is a progressive temporal blockchain approach that addresses
the aforementioned problems (SeFra). Specifically, SeFra's foundational components, the Context-based
Merkle Tree (CBMT) and Enhancement of access control, confidentiality, and integrity checking
procedures is the goal of context-based access control (CBAC).. Additionally, the framework allows for
Interoperability and handles scalability issues with the aid of temporal properties, HL7 specifications, and
117
ISBN: 979-8-3503-4821-7
Interplanetary File System for data management (IPFS). The results of the study demonstrate that the
proposed framework SeRfas performance in securely managing the Personalised Micro Ledger (PML) is
affected by the temporal shadow concepts and context-based properties. Additionally, it helps record health
records as text with temporal properties and share them with colleagues according to the rules set out in a
context-aware smart contract. Importantly, the framework may be able to solve the problems of data silos
and supply unchangeable, secure transactions in the healthcare sector.Electronic health records (EHR) have
been linked to medical inconsistencies and worries about privacy and security. Although it is true that lack
of standardisation is not the only difficulty in electronic health records (EHR), it is arguably the most
significant. To solve the problems of interoperability and privacy, blockchain-based decentralised online
ledgers have already been proposed and implemented. Nevertheless, there are still a number of other
obstacles to overcome. Main technical obstacles include scalability, usability, and accessibility. The first
technical difficulty is the need to keep medical records in their various formats off-chain while providing
secure access control to them on-chain. The second difficulty is providing an adequate description of who
owns what data and how that data is shared during times of medical emergency. Accessibility remains a
problem, however, as formal verification of smart contracts takes a long time relative to processing large
batches of transactions.
Keywords: Temporal Blockchain, Hyperledger fabric framework, PRE, PML.
PAPER:CS850
Exploring various aspects in diagnosing Autism Spectrum Disorder (ASD)
K. Gayatri,N. Kundana Sri Bhanu,Ch. Lohind Sai Durga,T. Poorna Sai Neelesh,Srithar S,Praveen
Tumuluru,
[email protected]
Abstract— Background: Autism Spectrum Disorder, ASD is a neurological disorder that impairs an
individual's behavioural development, making interaction and communication skills challenging. The word
"spectrum" implies that every person is unique and has a different collection of features that distinguishes
them from others. There has not been a quantifiable method for diagnosing or treating autism, let alone
tracking the effectiveness of treatment, due to the vast range of aspects and lack of clarity surrounding the
causes of the disorder. As part of our research, the works concerning the detection of ASD have been
reviewed and reached to a clear understanding of how different sets of algorithms work together and
categorised those combinations accordingly. Methods: The literature search for our study was done using
keywords: Machine Learning algorithms, Autism, ASD, Deep Learning Methods, Autism Spectrum
Disorder, Artificial Neural Networks, Image processing, Speech processing. Results: To simplify the
detection of ASD, the study's findings were organised and presented in three major classifications:
extracting characteristics from (a) behavioural analysis (b) image processing (c) speech processing.
Conclusion: To perform a comparative analysis of all three aspects and find an efficient model or algorithm
in each category for detecting autism with the most accurate findings feasible after filtering the range of
results across all aspects.
Keywords: Autism, Machine Learning, ASD, Deep Learning, Speech processing, Image processing.
118
ISBN: 979-8-3503-4821-7
PAPER:CS852
Identification of Leaf Disorders Using Machine Learning Techniques
P.Manogyna Sai,V.SushmaSri,V.Hima Sailu,U.Pradeepthi,Dr.M.Kavitha,Dr.S.Kavitha
[email protected]
Abstract— A nation’s advancement depends on the Agriculture Sector. In this article we will be discussing
techniques that are used to detect the diseases and damages caused to plant leaves. Detection of various
kinds of diseases in plants helps farmers to produce good quality yield. This kind of evolution helps farmers
to detect the diseased plants and resolve the damage in the starting stage itself before causing huge damage
to healthy agriculture fields. We are using a CNN based approach to detect the problem. CNN is an
architecture that is built on networks mainly focuses on Deep learning design and more precisely used for
identification or recognition of images and gives us more accurate and precise results. As soon as we notice
the underlying diseases in plants the loss can be scaled down and can be beneficial to farmers. The current
work demonstrates how well the CNN model can be used to identify the different disorders in plant leaves.
In this scenario the images were captured from the field areas, and some are taken from the pre-existing
data set. Moreover, CNN we are comparing our model with KNN (K-Nearest Neighbors) is a paradigm of
supervised learning which is utilised for classification which classifies or anticipates how a specified data
point will be classified using near point. Later this data set was implemented using Machine Learning
techniques. Through such a comparison of the precision of images categorization using CNN and KNN. It
is evident that when we employ CNN models rather than KNN for image classification, we receive results
that are more precise.
Keywords— Image Classification, Deep Learning, Machine Learning, Convolutional Neural Network
PAPER:CS853
Comparative Analysis for Prediction and Classification of Mental Health Issues and Challenges
Using Hybrid Learning Techniques
Nagaraj P,Sudheer Kumar. E ,Aishwarya R,Arun Kumar M,Ishwarya Lakshmi S,Neyashree M
[email protected]
Abstract— Mental health conditions including sadness, anxiety, and sleep deprivation own up to the
emotional stress in young children, teens, and also in adults. It affects how a person thinks, ponders, feels,
or responds to a certain circumstance or situation. Only if one has both good physical health and mental
health, can an individual work productively and reach their full potential. Mental health is very important
at every stage of life, from childhood to adulthood. We gathered information from online datasets that were
readily available. For better prediction, the data has been label-encoded. To obtain labels, the data is
subjected to several machine-learning approaches. The model that will be developed to forecast a person's
mental health will then be based on these categorised labels. Working-class individuals over the age of 18
are our target market. The model will be implemented into a website when it is created so that it may
forecast the outcome based on the information provided by the user.
Keywords— Prediction, Classification, Hybrid Learning Algorithms, Machine Learning, Stress,
Depression
119
ISBN: 979-8-3503-4821-7
PAPER:CS857
Squad Help – An Emotion based Music Player
Ms.S.Meera ,Ms.R.Sharmika Sree,Chandrasekaran D K,Sathish N,Balakrishnan K.
meera.cse @ sairam.edu.in
Abstract—In the modern era, emotion detection through facial expression gained massive attention due to
the active works in this field. For the purpose of facial emotion detection a mixture of algorithms and
applications were projected and implemented. A music player is created based on facial emotion of the
user. This music player helps the user to automatically play songs according to the emotions of the user.
The emotions are recognized using Principal Component Analysis (PCA) and Support Vector Machine
(SVM) algorithms. The webcam captures the image of the user then reduces the multi dimensional data of
face using PCA algorithm and then classifies it with a SVM classifier to arrive at the mood of the user with
better accuracy. After emotion is identified, songs relative to the emotion will be played from predefined
directories.
Keywords – Facial expression , emotion detection , PCA, SVM
PAPER:CS860
Soldier's Position Tracking & Health Monitoring Optimization Model Using Biosensors
S. Nirmala Sugirtha Rajini,K. Anuradha ,T. Kirubadevi,T. Chandrakala,Veeramanickam M.R.M,
Raja Marappan,N. Bharathiraja,S. Bhaskaran
[email protected]
Abstract— The largest withstanding army in India is the military force. The soldiers in the army have the
most significant responsibility to safeguard the life of the people besides monitoring their health. Health
monitoring is challenging in varying climatic factors under difficult situations. It is essential to monitor
their blood pressure, heartbeat, and Oxygen (O2) level. This research develops a new system to monitor
the soldier's health using the internet of technology (IoT) and also tracks the location of the soldiers for
proceeding with the other process immediately according to the situation. The proposed system consists of
two basic units – the wearable device and the control unit. The soldier holds the wearable device, which is
implemented with a temperature sensor, O2 level indicator, and heartbeat sensor. This unit manages and
observes the soldier's wellness embedded with the global positioning system (GPS) technology to track the
location of the soldiers, which aids in the transmission of data to another unit in the control room. This
research utilises different sensors like a temperature sensor to recognize temperature, a sound sensor to
distinguish sound created, a PIR sensor to identify individuals, a heartbeat sensor to recognize heartbeat,
and an Oxygen sensor to distinguish oxygen level. The multi-layer inference system conserves the device's
battery power with 98.5% data reduction compared to other well-known methods. GSM technique is
additionally utilised for sending messages to higher authorities.
Keywords – position tracking; health monitoring; sensors; IoT; arduino controller; multi-layer; oxygen
level; heartbeat sensor; temperature sensor; battery power
PAPER:CS863
Smart Metering Gauge System
B. Uma Maheswari,J. Sylviya, J. Ramya, M. P Rajakumar,
[email protected]
Abstract—In recent days IoT plays a vital role and provides most promising results in the transformation
of the operations in many domains such as manufacturing various products, medical systems and
transportation etc. We can use IoT technologies to purify the water which is essential in this current polluted
120
ISBN: 979-8-3503-4821-7
environment. Even though there are many techniques to purify the water, Reverse Osmosis is one of the
best methods which is used in most of the places. The membranes in the Reverse Osmosis filtration system
remove impurities based on their size and shape. Some of the impurities, which are larger than the water
molecules, are not filtered by this system. This is best suited for industrial usage. The IoT enabled Smart
Water Purifier includes additional features such as an inbuilt TDS, PTC thermistor, Wi-Fi based leak
detector. The above-mentioned components are connected to the Microcontroller that is in turn connected
to the GSM module that works under Wi-Fi and transfers information or alerts the user in their mobile.
Benefits are that RO purifier is equipped to filter toxins such as lead, mercury, arsenic, chlorine, and
fluoride. These chemicals may cause diseases. IoT enabled RO water purifier systems consume less energy
as compared to other water purifiers and the power consumption is also very less. RO purification improves
the taste of water. The maintenance of an IoT enabled RO filter is extremely simple, it requires service only
twice in a year.
Keywords- Internet of thing, Total dissolved Solids, Arduino, Leak Detection, Temperature Sensor,
Reverse Osmosis
PAPER:CS865
SPEECH EMOTION RECOGNITION USING DEEP LEARNING
Kartikeya Srinivas Chintalapudi,Venkata Saroj Kushwanth Muvvala ,Irfan Ali Khan
Patan,SuryaKanth V Gangashetty,Harsha Vardhan Sontineni,Akhilesh Kumar Dubey
[email protected]
Abstract— Speech emotion recognition is the task of automatically detecting the emotional state of a
speaker from their spoken words. It is a growing area of research that has applications in various fields
such as human computer interaction, education, and psychology. There are several approaches to speech
emotion recognition, including the use of machine learning algorithms, which can be trained on large
datasets of annotated speech samples to recognize patterns associated with different emotions. Other
approaches include the use of linguistic features, prosodic features, and physiological signals such as facial
expressions and heart rate. One challenge in speech emotion recognition is the variability in the expression
of emotions across individuals and cultural groups. Another challenge is the need to accurately identify the
underlying emotion, as opposed to simply recognizing the presence or absence of an emotion. Overall,
speech emotion recognition has the potential to improve communication between humans and machines,
and to provide insights into the emotional states of individuals.
Keywords- Deep Learning, Emotion Recognition, CNN
PAPER:CS867
Effective Approaches Of E-Commerce Product Recommendations
M Lokesh N S P S Sai ,Sai Saran Tej R,Vyshnavi G,Vijay Kumar Burugari ,Sai Sree B ,Naresh
Vurukonda,
[email protected]
Abstract— One of the well-known applications of recommender systems is the creation of playlists for
audio-visual services, online retailers that recommend products, open online content recommenders as well
as content recommendations for social networking sites. These programs can be configured to work an
individual input, such as numerous inputs that may be discovered on both platforms, including news, books,
and searches. Additional factors well-known recommendation systems for subjects such as eateries and
dating websites. Journal articles, specialists, colleagues as well as financial services are all researched by
recommendation systems. Therefore, this recommendation method, which directly compares users and
121
ISBN: 979-8-3503-4821-7
products, cannot be used with a collaborative filtering model. Content acquisition and quantitative analysis
form the basis of a content-based algorithm. Many current recommender systems that use content make
recommendations depending on the analysis of textual information, as research on collecting and filtering
textual information is advanced.
Keywords— Recommender system, Shopping, E-Commerce, Similarity, Rate of product
PAPER:CS868
Mobile Charging System Based on Coin Detection
Dr.A.Vijayaraj, Eluri Venkata Pavan Kumar, Tunuguntla Gokul Akash, Kaza Venkata Siva
Yaswanth , Sure Teja, Subba Reddy Chavva
[email protected]
Abstract: The demand for mobile phones is increasing day by day. Modern civilization would not function
without cell phones, which are both the main source of entertainment and a lifeline for those in need.
However, serious battery problems in mobile phones continue to be a serious problem for manufacturers.
The given document describes an effective, affordable and ready-to-use solution for individuals to deal
with this problem. The purpose of this paper is to install these vending machine-like facilities so that
mobile phones can be charged very cheaply. This solution can be used in busy areas such as airports,
markets, conference rooms and rest stops.
Keywords: Modern ,Battery, Mobile phones, Vending machine
PAPER:CS869
New Recommender System for Online Courses Using Knowledge Graph Modelling
Bhaskaran S. M Vinoth Kumar,Dr. N. Bharathiraja,Ravindhar N V,Pradeepa K,Raja Marappan
[email protected]
Abstract— Nowadays, customers can get recommendations for products and services that fit their
anticipated needs in the context of a large body of data. Advances in this area provide novel concepts and
enable technology for use in e-learning settings. The existing recommendation systems typically focus on
improving the item side, but they don't consider the user characteristics in the recommendation. Thus,
they're ineffective and hence not a good fit for a virtual classroom setting. This research focuses on applying
semantics-based knowledge graph modelling with collaborative filtering (CF) to solve these issues. This
proposed framework applies the users' static features to aid in cold start issues for potential new users. The
proposed framework significantly outperforms the existing models and helps maintain acceptable
performance using better user-item interactions.
Keywords-recommendation systems; knowledge graph; collaborative filtering; semantic modelling; virtual

classroom; graph modelling; online learning
PAPER:CS870
Multi-Plant and Multi-Crop Leaf Disease Detection and Classification using Deep Neural Networks,
Machine Learning, Image Processing with Precision Agriculture- A Review
Yogesh H. Bhosale, Shrinivas R. Zanwar, Syed S. Ali, Nagesh S. Vaidya, Rajesh A. Auti, and Deepak
Kumar H. Patil
[email protected]
Abstract- Globally, more than 19,000 fungi are reported to infect agricultural crops with diseases. As the
supplier of human energy, crops are seen as being significant. Plant diseases can harm leaves at any point
during planting and harvest, greatly reducing crop productivity and the general market’s financial worth.
122
ISBN: 979-8-3503-4821-7
Consequently, the early diagnosis of leaf disease is crucial in farmlands. Agriculture profitability is a key
factor in economic growth. This is among the causes why plant disease identification is crucial in the
farming sector, as the presence of illness in plants is extremely common. If necessary precautions aren't
followed in these regions, plants suffer major consequences, which impact the grade, volume, or production
of the corresponding products. For example, the United States has pine trees that are susceptible to a
dangerous illness called small-leaf disease and the backbone of the Indian economy is crop plants. It is
advantageous to diagnose plant diseases (Black Spot, other leaf spots, powdery mildew, downy mildew,
blight, and canker) using an automated method since it lessens the amount of manpower required to
maintain megafarms of crops and does so at an incredibly preliminary phase(when they appear on plant
leaves). The computerized identification and classification of plant leaf diseases using an imagery
segmented system is presented in this work. It also includes an overview of various disease categorization
methods that can be applied to the identification of plant leaf diseases. In order to detect disorders in diverse
plant leaves, this study provides a review of diverse plant diseases and several classifying algorithms in
deep machine learning.
Keywords- Plant disease, Leaf disease, Plant leaf disease detection, Plant leaf disease classification, Deep
Convolutional Neural Network (CNN), Deep machine learning (DML)
PAPER:CS880
A Survey on Deep Learning based Video Surveillance Framework
Abdul Saleem L,Dr. E Venkateswara Reddy .
[email protected]
Abstract—Technologies that are used in automated video surveillance give the capacity of automatically
identifying security breaches or other potentially dangerous occurrences taking place inside the field of
view of the cameras. These technologies have applications in the areas of surveillance as well as the
detection of intrusions along perimeters. Recognizing anomalous behaviour in crowded environments in a
quick and efficient manner using technology is a very successful method for boosting public safety. Several
different automated and real-time surveillance technologies for use in security applications are discussed
in this study. The fact that public locations cannot be manually monitored is the most important factor in
determining their level of safety and security. The algorithms for strange behaviour have made efforts to
increase their effectiveness, resistance to pixel occlusion, generalizability, computational cost, and
execution speed. In a manner analogous to the current state of the art in anomalous behaviour identification
in crowded settings, researchers broadly divided techniques into distinct categories such as tracking,
classification. It has been discovered that hybrid learning approaches and deep learning methods provide
more satisfying outcomes during the categorization stage.
Keywords-video surveillance, Web of Things (WoT), deep learning, behaviour analysis.
PAPER:CS881
Predictive Analytics based Modeling of the purchase intention of electric vehicles, and understanding
the drivers and risks in their adoption for the people of Tamil Nadu in India
Gaurav Nagpal ,Ankita Nagpal,Naga Vamsi Krishna Jasti,
[email protected]
Abstract— While the adoption of electric vehicles by the World is very important to address the issues of
climate change, the rate of adoption of EVs is substantially low due to some of the barriers in their adoption
by the population. While several studies have been done in the past using different theories to measure the
perception of people towards the electric vehicles in the context of several countries including India, there
is no such study for a specific Indian state. Since India is the land of diversity where people with different
123
ISBN: 979-8-3503-4821-7
lifestyles and values co-exist, the state-specific studies can generate more insights for the policy makers,
manufacturers and marketers of EVs. Therefore, this study measures the perception of people of Tamil
Nadu, one of the prominent automotive manufacturing hubs in the country, towards the adoption of electric
vehicles through a structured primary survey. A logistic regression model has also been developed to
measure the purchase intention of the potential adopters. The important findings of the study and
managerial implications have also been discussed.
Keywords— virtual work, flexible work, working hours, work-life balance, work practice profiles, conjoint
analysis
PAPER:CS887
Blockchain Based Crowdfunding
Ms. Geetha S K ,Mr. NaveenKumaran R ,Mr. Kaushik Selvaraju ,Mr. Kishore C ,Mr. Nagha
Rathish A
[email protected]
Abstract— Crowdfunding is a method of online fund-raising process that was initially developed for
public members to make modest contributions to support the projects of creative individuals. Crowdfunding
uses blockchain technology to offer smart contracts for users. This allows us to offer crowdfunding in a
secure, transparent, and safe manner. The task of this work is to provide interactive forms for campaign
development and financial contributions. Both campaign makers and donors may develop and support the
campaigns by viewing or submitting requests for approval and fulfilling requests using this system. In
addition, the donor may be able to see the progress of the funds they provide. All transactions will be
recorded on the blockchain and stored as blocks. It is alluring to use smart contracts in blockchain. Without
the aid of a trustworthy third party, a blockchain-based agreement must be negotiated, carried out, and
enforced amongst unreliable participants. It is essential to develop executable code that runs on the
blockchain. Blockchain was initially primarily used as the basis for cryptocurrencies, but in recent years,
it has expanded to various industries. Blockchain is anticipated to be the most widely used technology as a
green way to conduct internet transactions. One application area for blockchain technology is crowdfunding
websites. The biggest problem with today's global crowdfunding market is that campaigns are no longer
under strict control, and some crowd-investment efforts have proven fake. By utilizing Ethereum smart
contracts on the crowdfunding site, this work aims to allay these worries by assuring that the initiatives
may be added within the designated time limit, eliminating fraud, and allowing the contracts to be fully
mechanically performed.
Keywords—Blockchain,Smartcontract,Ethereum, Crowdfunding, Cryptocurrency
PAPER:CS897
AI Powered Smart IOT Helmet for the Early Detection of Covid-19 and Its Variants
Mohanaprakash T A ,V.Divya ,S. Saranya, Naveen , S.Prakaash ,T.Sunitha
[email protected]
Abstract – The most serious threat for the year 2019 has been named as COVID-19, and it has already had
a big effect on people's lives and the world economy. The world has been terrified by this virus' horrifying
repercussions. Both rich and developing nations around the world have been devastated by this deadly
virus. The health industry is also subject to a significant burden and expense due to COVID-19. With
today's technology, continuous monitoring could save health care expenditures while also improving
control over people's healthy behaviours, epidemic management, and prevention. Furthermore, the majority
of monitoring studies that have been conducted so far have concentrated on the patch and have few
instruments for detecting COVID-19 and its modifications, such as the propagation of the Omicron virus.
124
ISBN: 979-8-3503-4821-7
Since a thermal and pi module camera with occluded sensors can continually monitor certain measuring
parameters, the main goal of this research is to develop an adaptive monitoring system and model for a
smart artificial intelligence (AI) helmet. Real-time calibration and incorporation of body temperature and
facial identifying features. Additionally, an AI system would be used to support the proposed
functionalities.
Keywords – Artificial Intelligence, COVID-19, Monitoring Technology, Thermal body temperature, Smart
Helmet.
PAPER:CS899
Vehicular Mishap Analysis Based On Sensor
SHIVA KUMAR.P ,MURUGANANDAM.S
[email protected]
Abstract—Communication technology advancements have a profound impact on the automotive sector.
On-Board Diagnostics (OBD) devices are standard in many modern passenger vehicles and are available
for retrofitting in others. A connected car is one that is equipped with several sensors and can relay vehicle
data to improve user convenience, safety, and fuel economy. The Internet of Things (IoT) has the ability to
analyse the vast amounts of driving data generated in real time by sensors in connected vehicles in order to
provide useful inference or advice to drivers (IoT). The availability of such a wealth of data inspired the
creation of a data- supported solution to promote cost-effective and risk free vehicular transportation.
Keywords: Car, OBD, sensor, Fuel, safety.
PAPER:CS900
Stacked Bidirectional-LSTM Network for FakeNews Detection on Twitter Data
Paladugula DineshBabu,Taneem Kowsar Shaik,M. D. V. Sampath Kumar,P. M. Ashok Kumar,
Boyina Vishnu Vardhan Rao,R. Shankar
[email protected]
Abstract—This work suggests a computational approach incorporating deep-learning methods to combat
false information. The goal is to raise the false information detection scores by utilizing vectorizers, namely
Word2Vec vectorization and the elimination of English stop words. To produce a more precise forecast of
false information, our proposed model also makes use of a variety of classifiers, including Stacked
Bidirectional-Long Short-Term Memory (Stacked Bi-LSTM), Support Vector Machine (SVM), Decision
Tree, etc. By padding each sequence to its maximum length, we offer the network normalized qualities that
can affect how effectively and accurately it operates. The network then uses a Stacked Bi-LSTM to interpret
this data. Using the fake news data set, we made predictions regarding the validity of news. Our testing has
determined that the proposed solution is more accurate than other machine-learning models.
Keywords—Bi-LSTM, Text Preprocessing, Word2Vec
PAPER:CS901
Mid-Air Gesture Based Multi-Finger Control System For Paralyzed Patients Using Leap Motion
SNEHA. K
[email protected]
Abstract—Social security, mobile technology, system security, art and culture, sign language recognition,
defense technologies, and soon have all recently paid a great deal of attention to image processing, and
specifically Air Gesture Recognition (HGR). One of the most important uses of Air movements is in sign
language, and this study has implications for resolving issues with this vital communication modality.
125
ISBN: 979-8-3503-4821-7
However, its lack of universal comprehension undermines the entire sign language environment. The
computer's ability to recognise Air gestures is being put to good use in this area. Therefore, it is necessary
to develop an adequate Air gesture detection system for use in robotics, gaming and virtual reality, human
computer interaction, sharing information contained in historical artefacts, etc. Current methods for
recognising Air gestures typically rely on either a sensor based or a Vision-based approach. Sensor-based
approaches, such as gloves, Inertial Measurement Unit (IMU), Electromyography (EMG), Wi-Fi, etc.,
record the location and motion of the Air, while vision-based approaches acquire the Air gesture using a
camera, webcam, etc. The sensor based method is more efficient because it eliminates the need for pre-
processing and segmentation; nevertheless, it is prohibitively expensive to set up a dedicated laboratory for
this purpose, and it can be unpleasant for users to wear bulky sensors for extended periods of time. Because
they encode texture and colour aspects of Air gesture for recognition, vision-based techniques are cost-
effective. However, there are a number of obstacles that must be overcome in order to create a vision-based
Air gesture detection system. The study's overarching objective is to enhance existing approaches to
problem-solving and system efficiency in order to design a system that can successfully navigate these
obstacles. Pre-processing, feature Extraction, and classification are the three main building blocks of any
Air gesture detection system. In this study, we provide LMS (Leap Motion sensor) , a feature-based
descriptor that was built from the ground up. The static Air gesture identification problem is addressed by
suggesting an Extended Radial mean response pattern. When compared to state-of-the-art Aircrafted
methods, the suggested descriptor improves recognition system robustness to noise and light fluctuations.
Keywords : LMS, Air gesture, Classification, image processing.
PAPER:CS904
Incorporating of Security Methods into the Software Development Lifecycle Process (SDLC)
SaiVishal Pediredla ,Bhumula Vasanthi ,Dr. Padmanaban k,Batchu Radhika ,Dr. A.K Velmurugan
[email protected]
Abstract—As software usage increases and updates, Software development requires several essential
elements, including industry security. As vulnerabilities are exploited, we've witnessed an upsurge in the
demand for securities. There are many various reasons why security is not effectively maintained in
software development processes, thus we have seen some approaches like traditional security, CI/CD, and
robust security to make sure security is implemented right from the start rather than after process
development. The various tools in this study were used to discuss various security implementation methods,
and this discussion may be useful for examining various software development processes.
Keywords— Devsecops, security robotization, CI/CD
PAPER:CS906
PREDICTION OF PHISHING CYBER ATTACKS USING HYBRID DEEP LEARNING
TECHNIQUES
Manoj Kumar S ,Nandhini D
[email protected]
Abstract— Everyone today relies heavily on the internet. Everybody used the internet to do online
transactions and other online activities including banking, booking, and recharging. Phishing is a form of
social engineering assault that is frequently employed to obtain user information, such as login credentials
and credit card details. Automatic URL detection is more crucial as the Internet expands if end users are to
receive timely security. The suggested system is designed to deliver an efficient and adaptable malicious
URL detection system with a comprehensive collection of properties that represent numerous
characteristics of phishing websites and their hosting platforms, including elements that are challenging for
126
ISBN: 979-8-3503-4821-7
criminals to forge. This system will assist in identifying the quantity of questionable emails and add a new
layer of security to today's unstable global environment. When tested against a variety of common phishing
datasets, the suggested approach shows promising results.
Keywords— Prediction, phishing prediction, deep learning, hybrid learning, Gradient Boosting Classifier,
cyber security, secure websites, secure browsing, cyber attack prediction
PAPER : CS907
Methodical Tamil Character Recognition Using Fabricated CNN Model
Bhavani M, Ravikumar S, Dr. S. Prithi, Arockia Raj Y, Babu Rajendiran
[email protected]
ABSTRACT--Recognizing text is one of the most challenging problems in computer vision. To put it
simply, handwritten character recognition is the process of recognising text in photographs, papers, and
other media and converting it into a form that a computer can read. You can solve this issue by employing
the image classification method. We will be able to recognise the user's typed Tamil characters in this
paper. We designed a brand new model, MVGG16, to handle character recognition and classification. Its
basic structure is analogous to that of VGG16, but we added a few more layers to improve accuracy. So
far, 156 Tamil characters have been learnt, with 350 images each class. In experimental settings, this model
achieves an accuracy of 91.8%. As an input method, this work makes use of a whiteboard where the user
can scribble the Tamil character to be recognised. The character will be sent for verification after the user
has written it and the image has been exported. Once the model identifies the character, it will show it to
the user.
Keywords: Handwriting Character recognition, Image Classification, MVGG16, VGG16.
PAPER : CS909
Applicable Techniques for Image Steganography: A Survey
Dr. Ambika, Dr. Virupakshappa
[email protected]
ABSTRACT-- Steganography and cryptography are two different technical approaches preferred to meet
data security over the communication network. Due to the increase of illegal hacking, information security
needs alerts in high volume. With the growth of communications based on the internet, computer networks
need specialized security which would be available to secure incoming broadcast messages from any threat
producing hacker affecting data privacy and also ban access illegally. The process of transferring the secret
data without letting the attackers know. It is the method of transporting the secret message without grabbing
the notice of attackers. ‘Stegano’ and ‘graphy’ are two Greek words that mean ‘Covered’ and ‘Writing’
respectively it is beneficial for hiding the message, thus protecting it. In this paper, we've discussed various
steganography techniques, methodologies, along with applications and categorization.
Keywords: Digital image, information hiding, steganography, Security, Cryptography.
PAPER : CS913
Driver State Monitoring System Using AI
Dr Saranya N, Mrs. Priyanka V, Akhila B, Harini T, Hemalatha M A, Kaneshka Sre R S
[email protected]
ABSTRACT-- Artificial intelligence (AI) is a rapidly expanding field that the world needs. It is frequently
used in scenarios where a group of intelligent robots utilize their intelligence to weaken human power.
127
ISBN: 979-8-3503-4821-7
Whether we realize it or not, artificial intelligence plays a vital role in our daily lives. Artificial intelligence
and machine learning have replicated human brain processes to offer the more what it wants and to help
society go technologically advanced with digital evolution. The state and emotion of a driver can be
predicted using machine learning techniques, which can then be utilized to deliver information that will
increase road safety. It uses artificial intelligence in some way. The face, a crucial bodily feature,
communicates a lot of information. Facial expressions, such as blinking and yawning more frequently than
usual, differ from those in a normal state when a driver is fatigued. Systems can automatically learn and
develop without being explicitly programmed thanks to artificial intelligence. Bio-indicators, a driver's
conduct while driving, and facial expressions can all be used to gauge a driver's health. We give an
exhaustive survey of driver sleepiness detection and alert systems in this study. We also discuss various
machine learning methods that are utilized to assess the driver's condition, including NumPy, the HOG
algorithm, the SVM algorithm, the HAAR-based cascade classifier, and OpenCV. Finally, we list the
difficulties that the present systems are facing and discuss the corresponding research potential.
Keywords: Driver, Camera, Fatigue, Fatigue detection, Deep learning, Face recognition, Image Processing,
Extraction, State monitoring, feature extraction, eye tracking, Artificial Intelligence, Drowsiness Detection,
Machine Learning, Drowsiness Detection, Eye Detection, Face Detection, OpenCV, Viola- jones, HAAR
cascade classifier, Aspect Ratio, Road Accident, Road safety, Pattern Analysis.
PAPER : CS914
PREDICTION OF LIVER DISEASE USING STATISTICAL MACHINE LEARNING METHODS
S. Manoj kumar , V.Anbu ,S.K Abhishek ,N Jeevanantham , H.Ashwin kumar
[email protected]
ABSTRACT-- In India, liver disease ranks as the tenth leading cause of death, causing 2.95 percent of all
deaths. According to the World Health Organization, liver disease is one of the main causes of death in
India. With approximately 10 lakh new cases being diagnosed annually, it has developed into a significant
threat in India. One of the most critical parts of automated disease detection and prediction is data mining.
Medical data is analyzed using data mining algorithms and methods. Disorders of the liver have increased
dramatically in recent years, and liver disease is now one of the leading causes of death in several nations.
The patient datasets are looked at so that classification models that can predict liver disease can be made.
The aforementioned study used feature implementation and comparative analysis to increase the accuracy
rate for liver patients over the course of three phases. On the existing datasets of liver patients derived from
the first process's data sets, the min-max normalization algorithm is applied. In the second stage of liver
dataset prediction, PSO feature selection is used to extract a subset (data) of liver patient datasets from all
normalized liver patient datasets that only contains significant attributes.
Keywords: Random Forest Classifier, SVM, Decision Tree Classifier
PAPER : CS915
SMART IRRIGATION SYSTEM FOR COIMBATORE DISTRICT
S. Manoj kumar , Keerthivasan E, Praveen M, Kiruba Sankar R
[email protected]
ABSTRACT-- In India, agriculture is crucial to the growth of the food industry. Agriculture in our nation
is reliant on monsoons, which are an insufficient source of water. Therefore, agriculture uses irrigation.
The Internet of Things (IoT) represents a turning point in technological development. IoT is crucial in many
industries, including agriculture, which could eventually provide food for billions of people. This study's
objective is to deal with this issue. Since the entire system is microcontroller-based and wireless
programmable, there is no need to worry about timing irrigation according to crop or soil conditions.
128
ISBN: 979-8-3503-4821-7
Sensors are used to measure soil properties such as soil moisture, temperature, and air moisture. The user
(farmer) controls decision-making by utilizing a microcontroller. The sensor data is transmitted wirelessly
to a server-based database. The watering will be automated after the field's temperature and moisture
content have dropped. The condition of the field is periodically updated via mobile notifications to the
farmer. This system will be more beneficial in locations with a shortage of water and will be effective in
meeting its needs
Keywords: Smart Irrigation, Sensors, Bluetooth communication, Android.
PAPER : CS916
MACHINE LEARNING BASED PLANT DISEASE DETECTION USING CONVOLUTIONAL
NEURAL NETWORK
Ch. Siva Gopi, Ch. V. S. Chakravarthi, K. Bhuvanesh, D. Mohan Rajesh, Dr. Mohammed Ali
Hussain, Dr. M. Kameswara rao
[email protected]
ABSTRACT-- Agriculture is the basic income for the majority of those living in India. Many people rely
on agriculture for a portion of annual income. Almost 60% of people depends on agriculture for their
livelihood. Farmers do not have a full knowledge on crops and their diseases. Many new diseases are
occurring to the yield, detection of those diseases can be helpful for the farmers. Identification of these
diseases plays the main role in to prevent the losses in the yield. Volatile and bio-photonic sensors are used
to detect the diseases occurred to the leaves and plants. Photonic sensors capture the images known as
image processing and detects the disease. Farmers also have various types of diseases and they hurt
themselves while they are in the field. Respiratory rate is the main problem farmers are facing. Monitoring
the periodic temperature variations of thermal images at nasal locations allows one to calculate respiratory
rate.. In this paper we have studied about various diseases occurring to the plants and detection of those
diseases by using image processing
Keywords: Disease detection, Image processing, Bio-photonic sensors, features extraction, classification,
symptoms.
PAPER : CS917
CROWDSOURCING FRAMEWORK FOR AGRICULTURAL POWERED BY AI
Ms.Geetha S K, Mr.Hariharan K, Mr.Gokul A, Mr.Boopathiraj B
[email protected]
ABSTRACT-- — Recently Blockchain has reached heights and depths to solve the problems like
Authenticity, Integrity, and Availability of data that raises in conventional technologies. Blockchain also
solve the problems that exist due to middleware services, by eliminating the need for middleware services
and implementing a keen sense of logic to validate and authenticate with the help of consensus theorems
and many other techniques. Crowdsourcing is the practice of getting services or collecting information
from a large group of people, and rewarding the group with an incentive. In simpler terms crowdsourcing
allows an individual to clarify or satisfy his/her need for specific knowledge from a community of people.
Crowdsourcing platform provides incentives or rewards to the participant who provides the best solution
to the problem. Crowdsourcing, when combined in the field of agriculture, can help farmers easily produce
solutions for other farmers, on their own past experiences. The farmers producing the best remedial solution
129
ISBN: 979-8-3503-4821-7
will be incentivized accordingly. When crowd-sourcing is combined with blockchain the rewarded farmer
or the participant will get the incentive without the interference of the middleware and brokers. When the
Dapp is powered by AI it assists in identifying the disease in the crops indirectly as a suggestion, but still,
it allows farmers to describe their diagnosis.
Keywords: Blockchain, Crowd Sourcing, Deep Learning, Distributed App, Cryptocurrency, React,
Solidity, Truffle, Ganache.
PAPER : CS923
The Comparative Study of CNN models for Breast Histopathological Image Classification
Chandana Mani R.K, Kamalakannan.J
[email protected]
ABSTRACT-- In Recent years, breast cancer has been considered as the prevalent type of cancer and
second deadliest cancer. Histopathological image analysis is the gold standard for cancer diagnosis. Since
CAD systems take less time compared to traditional approaches it is notable that automated cancer
classification will be helpful for reliable and accurate diagnosis. Also CAD models assisted with deep
learning techniques produce excellent results for classification of breast histopathological images.
Convolutional Neural Network (CNN) models have achieved significant progress in medical image
classification tasks including breast histology images. This paper makes a brief comparative study that
illustrates performance of CNNs such as VGG-16, ResNet-50 and AlexNet for breast histological image
classification. The study used BreakHis dataset for its model comparison. Further the paper illustrates
application of data augmentation methods for the dataset as it improves the model accuracy. Also we extend
our study to transfer learning methods and cross-check the results after using pre-trained models for breast
histology image classification.
Keywords: Histopathology, CNN, classification, data augmentation.
PAPER : CS927
Software Defined Network Framework & Routing Protocol Based on VANET Technology
Pamidimukkala Lahari, Ravella Srilatha, Omkar Eswar Chejarla, Ravuri Yogesh, R Shankar, P M
Ashok Kumar
[email protected]
ABSTRACT-- In recent years, different types of researchers have focused on the topic of vehicular ad hoc
networks (VANETs) because of the different services they provide. VANET is a subset of Mobile Ad-hoc
Network (MANET), which refers to a group of smart nodes. In present days, communication within the
moving vehicles takes place because of the improvement in technology. These smart vehicles communicate
within them or with a roadside unit to provide road safety, a more protected and effective driving
experience, and protection against intruders. Messages in VANET are sent using open wireless channels.
In recent years, several Vanet technologies are being developed to provide the better communication in
between the vehicles, to detect accidents and transforming that information to other vehicles through
packets. Even though vanet provides several services, there is an issue with this technology which can
cause low latency in communication or delay in transferring the message. To overcome all these issues
serval technologies have been emerging to provide the efficient communication between the moving
vehicles. In this paper, we have discussed about the architecture of the Vanet, present emerging
130
ISBN: 979-8-3503-4821-7
technologies and also about future challenges in Vanet. SDN (Software defined Network) framework is
used to provide the effective communication in terms of utilizing less delay. Fog Computing is one of the
emerging technologies in vanet, where it performs storage and processing functions without any
interruption of the third parties.
Keywords: Challenges, fault tolerance, Vanets, security, Routing Path, Software Defined Networks (SDN),
IOT
PAPER : CS934
COMPARATIVE STUDY OF LINEAR PRECODING TECHNIQUES
Kaviya, G.Geetha
[email protected]
ABSTRACT-- The primary focus of this research is a large - scale mimo circuit predating technique used
in the 5G core process. The main components of 5G are first explained, with premature technologies. The
average square error, maximum rate emission, zero compulsory and other premature techniques are later
carefully examined. Finally, the capacity of the primary approaches and the bit error rate is compared to
the simulation. This is the first article to explore every linear advance technique that is currently in deep
use in the best of our knowledge. Some people in the study illustrate the current challenges in building in
advance, where technical qualifications are often controlled by practical implementation barriers. We show
that a practical premature technique for massive MIMO is still a mystery, and we recommend continuing
research directions that require study.
Keywords: Maximum proportional exchange, linear pre speculation instruction, MIMO, zero force and
minimum average square error are words that describe 5G mobile technology devices
PAPER : CS936
Electronic Health Records & Data Management using Hyperledger fabric in Blockchain
Pooja V , Vijay K, Raghavi V, Bhuvaneswaran B, Eashaan Manohar
[email protected]
ABSTRACT-- Permissioned blockchain networks ensure safety and reliability by employing blockchain's
fundamental tenets, including the distributed ledger, smart contracts, and consensus. There is a potential
for data privacy violations in the existing condition of patient data interchange across healthcare
organizations if there is a lack of mutual trust. The introduction of blockchain technology has made it
possible for healthcare providers to confidently share patient medical records. To ensure the safe and
reliable transfer of patient information, a permission blockchain network built on the Hyperledger Fabric
framework is proposed and implemented between patents and medical institutions. Blockchain's effective
management of EHRs is made possible by a Hyper Ledger Fabric network. Results from live deployments
demonstrate that inaccuracies in data sharing between healthcare providers (such as hospitals, clinics, and
public health agencies) are eliminated by using a blockchain based on the Hyperledger fabric. The
accessibility of the network allows for the prompt, secure, and legally compliant transfer of patient
information across organizations. By using smart contracts as chain code, the blockchain has enabled
transparent and secure data exchanges while controlling access to private patient health information.
Results indicate Hyper Ledger Fabric provides excellent functionality, security, and transaction privacy.
Keywords: Blockchain, Hyper ledger, distributed ledger
131
ISBN: 979-8-3503-4821-7
PAPER : CS938
Obtrusion unmasking of Machine Learning-Based Analysis of Imbalanced Network Traffic
Dr. C. Siva Kumar , Himasree Kolla, B. Sravya, Grandhim Nikitha, Donthamsetty Lahya Sri
[email protected]
ABSTRACT-- Obtrusion unmasking involves identifying and detecting malicious or anomalous network
activity. Imbalanced network traffic refers to a scenario where the ratio of normal to abnormal traffic is
significantly skewed, with one type of traffic significantly outweighing the other. Machine learning and
deep learning techniques can be used to analyze network traffic and identify patterns and characteristics
that may indicate the presence of obtrusions or anomalies. These techniques can be applied to various types
of network data, including packets, flows, and logs, and can be trained on large datasets to improve their
accuracy and effectiveness. For the purpose of predicting the intrusion in imbalance network traffic, this
research employs a variety of categorization techniques. The classification algorithms are Random Forest,
Support Vector Machine, convolution neural network, and principal component analysis. The dataset,
which included several meteorological parameters, was obtained via the UCI repo. With a testing data ratio
of 70:30. Based on precision, accuracy, recall, f1-score, and execution time, the efficacy of the
classification algorithms was assessed
Keywords: Machine learning, Deep learning, Principal component analysis, Convolutional neural network,
Random Forest.
PAPER : CS939
Hybrid BiLSTM Network for improving crop pest classification
S.Pournima, C.Priyatharsini, G. Kirubasri, Manikandan J
[email protected]
ABSTRACT-- Insect infestations cause considerable monetary losses for the agricultural industries that
produce rice, millet, maize, legumes, sugarcane, garbanzo, and onions. It is essential to determine the
species of the bug as as quickly as possible in order to minimize the amount of money lost to it.
Nevertheless, farmers have a difficult time discriminating between the numerous types of agricultural
insects because they lack the experience and understanding to do so. In order to find a solution to this
problem, one option is to make use of convolutional neural networks, often known as CNNs, together with
other computer-based techniques. CNNs are useful for the identification of many images because of their
capacity to automatically learn features that are dependent on the information from the data themselves.
Within the scope of this study, we proposed a hybrid BiLSTM network as a potential solution to this
problem. A pretrained model and a layer of bidirectional long shortterm memories (BiLSTM) that takes
temporal information into consideration are included in the architecture that has been recommended.
Keywords: Long Short-Term Memory, Convolutional Neural Network, Deep Learning.Keywords-
component; formatting; style; styling; insert (key words).
PAPER : CS944
Empirical study of Human facial emotion recognition: A Deep learning
Sivakumar Depuru, S. Sivanantham, K Amala, M. Saktivel, V.Akshaya, P. Sai Kusuma
[email protected]
ABSTRACT-- A popular method for identifying human expressions is facial emotion recognition. Facial
emotion recognition identifies the non-verbal actions of humans. A more exhaustive and trustworthy
comprehension of the unique features of classifiers is revealed for researchers by the analysis of classifiers
gathered from recent publications. The pre-defined expressions like fear can be identified. An integrated
field of research on the capability of computers to perceive and comprehend sentimental situations of
human emotions. This paper explains the survey of facial emotions by using feature extraction,
132
ISBN: 979-8-3503-4821-7
classification of images, and data pre-processing. Based on geometry and appearance, there are two widely
used approaches that are frequently utilized within the written works for automatic facial emotion
recognition systems. Even though static images have been the subject of a lot of research, work is currently
being done to create new techniques that would be much simpler to compute and would use less memory
than older ones. The results of various facial recognition techniques are compared based on the different
algorithms. The datasets covered in this research include CK+, JAFFE, and Emotions with NLP dataset
and a few more. We anticipate that the emotion detection will be helpful in various applications, including
student predictive learning, and among others. In this survey paper CNN-LSTM based neural network was
proposed to improve the accuracy. Emotions with NLP dataset is used. The LSTM model has the capability
of using the data in series. We had got the accuracy of 98% by using CNN-LSTM.
Keywords: Natural Language Processing, Facial Emotoion Recognition, CNN-LSTM, Machine learning.
PAPER : CS945
A Hybrid Model Approach For Stock Market Price Prediction
Dr.P Dhanalakshmi, Penugonda Balaji, Samanu Dhanasekhar Reddy, Narre Diwakar, Sulake Vinay
[email protected]
ABSTRACT-- Customer satisfaction plays a most important role in predicting the stock prices in the
market. This paper shows how to identify the customer satisfaction based on numerical analysis of stock
prices and collecting tweets from twitter. Customer comments posted in social media websites like twitter
will be collected through snscrape module. Numerical data of stock prices were collected from yahoo
finance. Finally predicting the sentiment score using SVM and random forest algorithm, this will help in
future improvement of stock prices based on customer satisfaction.
Keywords: Sentiment, SVM, Random Forest, Twitter, Vader, Yahoo finance
PAPER : CS946
Multi-Object Detection and Tracking Using Machine Learning
Ayush Sahay, Keerti Vardhan Singh, Dr. Godwin Ponsam
ABSTRACT-- In today's information-based economy, data serves as a replacement for oil. There have
been shifts in how quickly and accurately benchmarks are measured as a result of efficient data. Industry
buzzwords Computer Vision (CV) and Artificial Intelligence (AI) perform the data processing, making the
improvement observable (AI). Two technologies have allowed for previously impossible endeavours, such
as the identification and tracking of objects for traffic surveillance systems. The need for an effective
algorithm to unearth concealed elements in images grows in tandem with the number of such features.
Single-object detection in the urban vehicle dataset is handled by a Convolutional Neural Network (CNN)
model, while multi-object detection in the COCO and KITTI datasets is handled by YOLOv3. Metrics are
used to evaluate and chart the performance of the models (mAP). On traffic surveillance video, we use
YOLOv3 and SORT to follow objects as they move between frames. This study argues that cutting-edge
networks like DarkNet deserve to be treated as special cases. We see effective detection and tracking on a
dataset of urban vehicles. The algorithms produce very reliable identifications that can be used in real time
for traffic applications.
Keywords: Object detection; object tracking; Computer Vision (CV); Urban Vehicle Dataset; Convolution-
Neural Network (CNN); You Look Only Once (YOLOv3); Artificial Intelligence (AI); Common objects
in Context (COCO);
133
ISBN: 979-8-3503-4821-7
PAPER : CS948
An Improvised Machine Learning Model KNN for Malware Detection and Classification
M. Raj Shekhar Rao,Deepanshu Yadav, Dr. V. Anbarasu
[email protected]
ABSTRACT-- The problem of network security has arisen as a key source of worry in today's linked
society. Sabotage and information extortion are among the most significant risks to an organization's
security. They include a broad spectrum of dangers such as huge intellectual property theft and software
attacks, as well as sabotage and information extortion. Inappropriate usage of protocols on a network may
potentially represent a security risk to a system or network. In network security, the use of data mining
tools to gather, alter, and analyze enormous amounts of data is vital. With the help of a range of data mining
technologies that are now accessible, it is possible to conduct analysis and forecasting of data and threats
across computer networks. This study makes an effort to forecast network security dangers by using a
variety of categorization methodologies. The classification techniques used in this study include the Naive
Bayes Classifier, the Decision Tree Classifier and the K Nearest Neighbors Classifier. It assesses the
efficacy of the categorization approaches described above in order to discover potential risks and
vulnerabilities. In order to analyze a dataset and extract information, Simple Machine Learning techniques
are utilized, such as in Distributed Denial of Service (DDoS) attacks (R2L), U2R attacks, and Probe Attacks
(U2R). The Naive Bayes Classifier, the Decision Tree Classifier, and the K Nearest neighbours method are
all examples of classification algorithms discussed in this section. For the purpose of comparing the
performance of the proposed machine learning algorithm technique with other approaches, the entropy
calculation and accuracy, recall, F1 Score, and entropy may be employed.
Keywords: Machine Learning, Computer networks, algorithm, entropy, accuracy, recall.
PAPER : CS951
Factors affecting gaming engines: Analytical Study
Mayank Tyagi,Reshu Agarwal, Amrita Rai
[email protected]
ABSTRACT-- The apparent evolution of machine learning in gaming is the application of machine
learning techniques. One major use can be the creation of much more realistic, smarter and responsive
characters in the games those which can learn new abilities from the various actions of the players and then
use these actions to counter-perform the tactics and the strategies as well as also produce many unscripted
responses when these actions are being observed by the in-game player actions. The methodology of
analytics is providing a better understanding of the gaming industry data and how it is going to affect a lot
of businesses in the future. Also, monetization models in gaming have been steadily shifting to lower
outspoken pricing all the way down to free-to-play balanced with increased in-game purchases and
microtransactions which give profit over time. still, this model of course requires both long-term player
engagement and desirable in-game purchases. Machine learning can help in the design of these products
and in deciding when and how to present them to the player.
Keywords: Gaming, Playing Characters, Data Analytics, Machine Learning, AI, Trends
134
ISBN: 979-8-3503-4821-7
PAPER : CS953
Predicting Flight Delay Based on Sentimental Analysis:Machine Learning
Mr.Shaik Salam, Kommireddy Dinesh Reddy, Thothireddy Yogiswar Reddy, Shaik Mohammad
Sohail, Vonka Darvesh, Oruganti Govinda Rao
[email protected]
ABSTRACT-- In the airline sector around the world, flight delay is the biggest issue. In the last 20 years,
the expansion of the airline business increased air traffic, which resulted in flight delays. This causes huge
losses in airline industry and also negatively impacts the environment. Therefore, it is crucial to avoid
delays and cancel flights. In the present work, the flight delay is predicted by using some regression models
like Linear Regression and Random Forest Regression, classification models like Logistic Regression and
Decision Trees and Sentiment Analysis. Then, based on accuracy, best model will be suggested. The
majority of earlier works are only appropriate for one route or airport. This study looks at a variety of
airlines as well as variables that could affect flight delays.
Keywords: Flight delay prediction, Linear Regression, Random Forest Regression, Logistic
Regression,Decision Trees and Sentiment Analysis.
PAPER : CS954
Forecasting Temperature and Rainfall for Crop Cultivation with Classification of Ground Water
Level
Dr.R.Thamilselvan, Dr.E.Gothai, M.K.Dharani, B.Aravinthan, S.Dharaneeshwaran, D.Jaya Suriyaa
[email protected]
ABSTRACT-- These days, wireless telecommunication networks offer a viable replacement for traditional
monitoring systems and rainfall measuring gear. Empirical computing techniques are less accurate
representations of real data because to the vast amounts of data connected to rainfall and telecommunication
networks. In order to analyse massive data and provide real measurements more accurately, deep learning
models are suggested. This project examined the outcomes of experimental measurements used to monitor
rainfall. This study's main objective is to demonstrate a neural network-based classification method for
rainfall data based on historical production data. Farmers can take the appropriate actions to gauge crop
yield in the upcoming season with the aid of classification based on the rainfall from prior years.
Understanding and evaluating crop output in the future can assist to protect food security and lessen the
effects of climate change. The ARMA (Auto Regressive Moving Average) approach is employed in this
work for the proposed work. For our country, data from the previous ten yearsare used to calculate rainfall
and ground water levels. In the proposed study, records in the ground water level data set are classified by
an ARIMA model, and what the model predicts for additional test record data sets. The brand-new design
help in researching past ground water levels and forecasting levels in the upcoming
Keywords: KNN Classification, Neural Network, ARMA Model, Crop Yield Prediction.
135
ISBN: 979-8-3503-4821-7
PAPER : CS958
Empirical Analysis on Crime Prediction using Machine Learning
R.Sathish Kumar, K.Nirmala Devi, N.Krishnamoorthy, N.P.Saravanan, P.Jayanthi, S.Karthi
[email protected]
ABSTRACT-- Criminal offence or Crime is an detrimental act that not only affects an individuals but
affects an community. In the current scenario crimes sequences happens every second in various places, in
various crime patterns, as the results the crime rate is increasing each growing places. It is very much
essential to determine the different factors for the possible occurrence of crimes and find the optimized
ways to reduce the crime. Crimes can be compartmented into multiple factors like Rape, Murder, Assault
and Theft and so on. Data were gathered from National Crime Records Bureau. This proposed work focuses
on how machine learning algorithms such as Naïve bayes, Random forest, Logistic regression, Decision
tree and Support vector machine can be outlined and analysed for reducing of crime rates in India. The
performance of proposed models are compared with various measures like accuracy, precision, recall and
f1 score along with the confusing matrix. Also the error value is analysed for various models with the help
of mean absolute error technique. As a result, the Naïve bayes model gives the best results of 98.94%
comparing with other.
Keywords: - Crime, Machine Learning , Support vector, Naïve Bayes, Logistic Regression..
PAPER : CS962
TRAFFIC SIGN DETECTION USING HYBRID NETWORK OF YOLO AND RESNET
Mr.S.Vinothkumar, Dr.S.Varadhaganapathy, Dr.S.Varadhaganapathy, Mr. S.Pradeev, Mr.
S.Pragatheeswaran, Mr. K.Sankara Annamalai
[email protected].
ABSTRACT-- Deep learning (DL) methods have outperformed machine learning (ML) and statistical
techniques in predicting road traffic. Neural networks serve as the foundation for deep learning algorithms.
In smart transport systems, the recognition and identification of traffic signs is a critical issue. Automated
traffic sign detection and recognition Driving is essential for achieving autonomous transportation.. There
are numerous traffic signs, and it takes time to train a good model. A TSR system can raise a driver's
awareness of the road situation and condition, potentially lowering traffic accidents. The YOLOv3 (You
Only Look Once, Version 3) algorithm, a real-time object identification technique with cutting-edge
features that detects individual items approach for detecting traffic flow, is what we suggest as a solution
to this problem. The YOLO algorithm uses features that have been learned to find objects. For more
accuracy we are combining two algorithms to make the prediction more efficient and accurate. ResNet is
used to extract features from the training data set and the classification is done by YOLO V3 algorithm.
Versions 1-3 of the YOLO machine learning algorithm were produced by the third version of the
methodology, which is an improved version of the original DL algorithm. Before classification and the
output of the optimised result, the image will undergo pre-processing to extract features. The proposed
method makes it easier for drivers to detect traffic signs and reduces the number of accidents on the road.
Experiment results show that the YOLO-v3 approach outperforms other previously existing algorithms in
terms of average accuracy
Keywords: Yolo, ResNet, Convolutional Neural Network, Pooling layer.
136
ISBN: 979-8-3503-4821-7
PAPER : CS967
HAND WRITTEN TEXT RECOGNITION AND GLOBAL TRANSLATION
Vijay K, Mukundh A, Pooja S, T Nithesh Aravind, Manoj Adhithya R
[email protected]
ABSTRACT-- Handwritten appreciation is the capability of a machine to recognise handwriting on paper,
in a photograph, on a digital writing device, etc. Among other places, financial institutions, corporate
settings, and manufacturing facilities are just a few of the potential applications for handwriting and
character recognition technologies. The key area of the research is to generate a method based on a
convolutional neural network that can recognise each character in a particular type format with accuracy.
The way a neural computer works defies conventional wisdom. The objective of training a neural computer
is to get it to either classify input data into one of several categories or to let the data evolve so that its most
desirable characteristics show up. Since neural computing is still a relatively new idea, its structural
elements are less well defined than those of other architectures. In order to convert the handwritten text
into digital text, this project categorises each handwritten word. To achieve this goal, we combined two
methods: character segmentation and direct word classification. In the first method, we train a model that
reliably categorises words according to their meaning using a Convolutional Neural Network (CNN) with
various topologies. In the latter, we create bounding boxes for each character consuming Long Short-Term
Memory (LSTM) networks with convolution. The classification by CNN of these segmented characters,
each word is subsequently reconstructed using the results of the classification and segmentation.
Additionally, we now have regionallybased worldwide translation that is adaptable for users and clients.
Keywords: Yolo, ResNet, Convolutional Neural Network, Pooling layer.
PAPER : CS971
Attempting to Perform Assessment for Plant Breeding Using Scored Entropy-based Choice
Hierarchie
KARTHIKEYAN N , LEKHAA TR , RAMYADEVI N , SUMATHI P
[email protected]
ABSTRACT-- Based on the increasing advancements in breeding tools immense breeding is possible
where voluminous information assessments has to be accomplished by software digitally. Here the goal is
to assess diversity within the gathered information using information extraction schemes. Based on the
features the assessment in plant reproduction is regarded as the actual process of categorization and the
scoring entropy based choice based hierarchical scheme is designed for performing the categorization. The
scheme employs chronological quality phenotype and assessing information for building choice based
hierarchies which are further employed for creating assessments for forthcoming cultivars based on their
quality phenotype. In order to illustrate the efficiency certain assessments were performed based on the
three varieties of green pea validation trials in terms of prior development (PD), average development (AD)
and green peas (TG). The work could be extended based on open breeders from immense number of
fundamental works with improvements in effectiveness of plant breeding.
Keywords: Plant Breeding, Chronological Quality Phenotype, Choice-based Hierarchies, Open Breeders
and Information Extraction
137
ISBN: 979-8-3503-4821-7
PAPER : CS976
Accident Prevention System using Machine Learning
Srihan Thokala, Murari Alli, Rohith Jakkani, Hariharan Shanmugasundaram
[email protected]
ABSTRACT-- Most of the Accidents Occur in these days are during the Night time. As per the Reports of
the Accident, Most of the Accidents are due to Improper vision of the Drivers during the Night time. In
order to minimize and prevent the Accidents we came with a Machine Learning Model Which Continuously
Monitors the Road in the Range of 20 Meters and Specifies if there are any people or Animals passing
across the Road. The Significance of the project is to help the innocent people who might lost their lives
due to the Accident without their intervention. The Model is also useful for the animals and passers. The
main purpose of the project is to detect strollers or animals like dog at night and dim light conditions. As
the light intensity during night is less, Even our human eye cannot detect a person. The existing system are
less effective due to the less accuracy of algorithms. The night-vision systems indeed work on mainly image
processing with assistance of camera and processing units
Keywords: night vision, machine learning, strollers
PAPER : CS985
Monitoring and accelerating plant growth using IoT and Hydroponics
Abhishek Kumar, Dr. Savaridassan P
[email protected]
ABSTRACT-- Growing plants without soil is known as hydroponics, and it is particularly helpful in places
with little space and bad weather. A novel method of growing plants without soil in an inert medium is
called hydroponics. By avoiding pesticides and other substances that degrade soil and land quality,
hydroponics enables farmers to conserve water while enhancing crop quality. Due to its high efficiency
and role as an environmentally benign substitute for soil culture, it has the ability to close the gap left by
India's low agricultural productivity. This farming process has leveraged several new technologies like
Internet of Things (IoT), making integration of intelligent management and real-time data monitoring and
collection systems possible. This farming process also allows for a greater degree of automation and
refinement in the control process of the hydroponic systems. Via this paper, we have discussed the
development of a hydroponic subsystem that leverages agents for cloud-based monitoring, data collection,
and storage for future analysis and heuristics. This paper presents an intuitive and low-cost approach to
monitoring and accelerating plant growth based on Hydroponics and IoT with tools such as automated
environment parameter monitoring and error rectification, sub-second streaming of telemetry from ground
units to cloud servers and allowing users to access the status of the plant or system of plants being monitored
as a plant health index from anywhere in the world, via a mobile application.
Keywords: -Internet of Things (IoT), nutrient acquisition, smart agriculture, hydroponics, cloud-based
applications
138
ISBN: 979-8-3503-4821-7
PAPER : CS987
Stock Price Trend Analysis and Prediction of Closing Price Using LSTM
Sparsh Vohra, Dr. Savaridassan.P
[email protected]
ABSTRACT-- One of the most challenging jobs in the realm of computing is stock market forecasting.
The forecast is influenced by a wide range of variables, including physiological vs. physical elements,
rational vs. illogical conduct, investor attitude, market rumors, etc. All of these factors come align to make
stock values unpredictable and exceedingly challenging to forecast accurately.ML approaches have the
power to disclose patterns and insights that we hadn't previously seen, and they can be leveraged to produce
forecasts that are incredibly accurate. We offer a framework to examine and forecast a company's future
growth utilizing the LSTM (Long Short-Term Memory) model and the net growth calculation technique
for businesses. Using variables like the opening price, closing price, and trade low, this may be anticipated.
Keywords: - LSTM, Machine Learning, Patterns, Close price, Open Price, Trade low, Trade High.
PAPER : CS994
Review Paper on Facial Recognition Techniques
Sujith Krishna Mulpuri, Damisetty Gayathri Lakshmi, Geethika Gudapati, K. Naga Lakshmi
Neelima, Dr. Thati Anuradha, Dr. Suneetha Bulla
[email protected]
ABSTRACT-- In this review paper, we focus on facial recognition techniques that use machine
learning(ML) algorithms to examine facial features and identify individuals. These techniques include
PCA, LDA, ICA, SVMs, and ANNs. Each of these methods has its own benefits and drawbacks, and the
best approach for a particular application will depend on factors such as the accuracy and speed required,
the type of data available, and the specific use case. In tzhis review paper, we provide an overview of these
techniques and discuss their capabilities and limitations, as well as the challenges and opportunities
associated with their use for facial recognition. We also consider the ethical and legal implications of facial
recognition and the need for responsible and transparent use of technology..
Keywords: - Linear Discriminant Analysis(LDA),PCA, Support Vector Machine(SVM), ANN,
Independent Component Analysis(ICA).
PAPER: CS1004
Secure and Proficient Provable Data Procurity With Privacy Protection in Cloud Storage
Dr.C.Siva Kumar, Lebaku Vasantha, Pacha Siroshini, Pullaiahgari Jagruthi, Nellore Pavithra,
Thumati Supriya
[email protected]
Abstract—Cloud technology is the distribution of physical servers in such processors, memory, networks,
connectivity, applications, statistics, as well as smartness only via Online platform so,in addition to allow
greater innovative technology, better versatile facilities, plus scalable enterprises.With this mechanism, the
data owner's maintenance and storage costs can be reduced. Simultaneously, when the data owner loses
procurity and physical authority over the data, a security threat arises. As a result, data inspection is required
to ensure data integrity over a server.However concern is now difficult for information owners because data
authority must have been validated thereby preserving confidentiality. In the process of labelling these
issues, this paradigm provides secure data acquisition as well as efficient privacy-preserving
verification.Additionally, it now supports multiple ownership, data complexities, also an unit identification.
A majority intriguing aspect for the process in which examiner will validate ownership of data having little
maintenance cost. The system demonstrates that a malware platform will present evidence that is adequate
139
ISBN: 979-8-3503-4821-7
to carry third-party auditors authentication of the user. The scheme's security is analysed and demonstrated
at this system to protect data privacy from third-party public auditing while making it impossible for cloud
service providers avoids saving the necessary blocks to ignore the result.So proposed scheme's best
intriguing feature is that it supports all of its crucial elements, such as blocking-free validation, security,
process monitoring, as well as information behaviour.
Keywords—SEPDP, Data owners, User, Storage in the cloud
PAPER: CS1005
Identification of Pneumonia Symptoms in Covid19 patients using Transfer Learning Approach
EBIN PM, ATHIRA B KAIMAL
[email protected]
Abstract— Over 1 million individuals were impacted globally by the COVID 19 epidemic, which also
claimed over 10 lakh lives. As a result of the Covid 19 infection, pneumonia might develop, putting the
patient in danger of serious illness or even death. Therefore, it is crucial to recognize the signs of pneumonia
and its existence in Covid 19 patients. The VGG16 architecture is a Deep Learning architecture that was
the first runner-up in the 2014 visual recognition challenge. The researchers are applying transfer-learning
to detect the presence of pneumonia in this case. Chest X-ray scans from kaggle, a publicly accessible open
dataset, served as the study's data set. The model's accuracy was 95.83%, and a comparison with various
other models was also presented.
Keywords—Covid19, Deep Learning, Transfer Learning, VGG16, X-Ray Images.
PAPER: CS1007
Vehicle Lane Detection for Accident Prevention and Smart Autodrive Using OpenCV
Manav Garg, Apeksha Sehrawat, Dr. Savaridassan.P
[email protected]
Abstract—Road accidents have been rising alarmingly, and the reason for this increment is because of
drivers being distracted. As a result, several technical advancements have lately been developed in the field
of road safety. There are many systems available that can alert drivers to various threats, including lane
departure, the potential for collisions, and numerous traffic signs. Through the use of lane detection
systems, which essentially assist by identifying the lane borders of the road and then alert the driver if he
moves to an improper lane, it is one way to go about this, moreover during the night times where chances
of less visibility and losing concentration by driver can occur would also be looked over. So the use of
multiple techniques can be enhanced by upscaling the images. The use of gamma correction is going to
help to increase contrast and upscale the images for better recognition of the lanes at night which are dark
and are hard to analyze due to less light.
Keywords— Lane Detection, Artificial Intelligence, Computer Vision, OpenCV, Gamma Correction
PAPER: CS1008
An IOT Framework based Automated Wireless Meter System for Monitoring, Billing and
Controlling Power Consumption
HARINI.K, Dr. V. ANBARASU
[email protected]
Abstract— The amount of power used in private houses nowadays is rising daily. Because we must
physically travel to the meter room to gather readings, verification and ongoing monitoring are highly
important tasks. Well, a consumer needs it since they will be charged based on how much power they use.
One of the biggest problems facing the globe today is how to manage its energy resources, particularly in
low-income developing nations like Malaysia where imports account for the majority of energy production,
placing a heavy burden on the national economy. Researchers and those in charge of making policy are
suffering because of this abundance. Over time, the domestic energy landscape has seen tremendous
transformation. Therefore, effective energy management is a critical problem that must be addressed to
help the economy move toward a higher development trajectory. Energy supply, demand, and pricing
challenges require a comprehensive approach to be addressed. The Smart Energy Meter is a tool used for
energy accounting and auditing, distribution transformer centers, and substation feeders that monitors
140
ISBN: 979-8-3503-4821-7
energy most precisely in industrial settings. Smart meters are used in solar, wind, and thermal power
facilities. Energy meters use parameters such as instant parameters, load profiles, daily profiles, and billing
profiles. Instant profile enables the assessment of voltage, current, phase angle, active power (KW), reactive
power (KVR), and perceived power in real time (KVA).
Keywords: Meters, Monitoring, Relays, GSM, Smart Energy Meter.
PAPER: CS1043
The Role of Artificial Intelligence in Education
Swathi Ganesan, Dr Nalinda Somasiri, Sangita Pokhrel
[email protected]
Abstract -- The paper depicts the design of the Intelligent Tutoring System for computing to assist students
learn Computing. The essential thought of this tutoring system is an orderly presentation into the idea of
computing. The system presents the subject of computing and manages naturally created problems for the
student to understand. The Tutoring System incorporates three fundamental model architectures: Domain
knowledge model architecture, student model architecture, and Pedagogical Module Architecture. Utilizing
the domain knowledge module architecture, the ITSs ought to have the option to take care of issues
concerning the academic module which is given to the students. The incorporation of cognitive science
concepts into ITSs culminated, is what currently considered an intelligent tutoring program. Such structures
provide knowledge of pedagogical methods and the knowledge of the student as a pedagogical module and
a student module, in addition to the information about the teaching content. One of the limitations that
influenced the development of ITS is the difference in time zone. ITS development exceptionally relied
upon the teamwork of the project team. The members of the team can identify and remove the bugs in codes
via common cooperation, but the time zone difference can affect the bug removal process because of
changing time plans.
Keywords -- Computer Assisted Instruction, iTutor, E-TCL expert tutoring system, meta model MDE stage,
ontology, semantic language.
PAPER: CS1044
Deep Learning Approaches for Accurate Sentiment Analysis of Online Consumer Feedback
Swathi Ganesan, Dr Nalinda Somasiri, Chandima Colombage
[email protected]
Abstract— Over the earlier time, a category of machine learning, called deep learning, has attained
significant achievements in several computer vision tasks such as image classification, object detection,
semantic segmentation, pattern recognition and image classification generation. Deep learning objectives
at finding various levels of dispersed representations, which have been proven to be discriminatively
effective in many tasks. Distributed statement depicts similar information highlights across different
adaptable and reliant layers. Each layer characterizes the data with a similar degree of exactness, however
adapted to the degree of scale. The implementation of deep learning techniques depends greatly on the
variety of data interpretation (or features) on which they are used. Artificial intelligence plans to understand
interpretations of information regularly by changing over it or isolating components as of it, which creates
it simpler to play out an undertaking like order or extrapolation.
Keywords—Sentiment, LSTM, CBOW, Custom Embedding Model, Word2Vec Model, Averaging Model,
GloVe Model
PAPER: CS1045
Exploring the Intersection of IoT and Blockchain: An Analysis of Security and Privacy Risks
Harishma Ashok, Swathi Ganesan, Dr Nalinda Somasiri
[email protected]
Abstract— Internet of things (IoT) is one of the leading technologies of transforming the internet into the
next level. Security of IoT is becoming a major challenge and its measures of stopping these insecurities
have been developing. Improving IoT security is important since IoT provides a great environment for
systems in the industrial world and other several applications. The challenges need to be addressed so that
consumer's data can be safe. Measures available are to improve the protection of access by unauthorized
persons, providing a safe way to transfer data, and ensuring synchronization of the system. Blockchain has
also provided a safe way for IoT devices, each device is given the public key, IoT devices scramble and
141
ISBN: 979-8-3503-4821-7
send the message dependent on the public key of the objective device so just the objective device can
unscramble the message.
Keywords— IoT, security, Blockchain
PAPER : EC104
UNMANNED PLANT IRRIGATION SYSTEM USING IOT
R Sundar Ram, Easwar G, Thenmozhi Karuppasamy, Avila Jayapalan
[email protected]
ABSTRACT-- Agriculture is an important asset to humanity because it relies on water to provide us with
food. This paper focusses on building a system which helps the user in monitoring the irrigation process
and inform the user in case of emergencies like smoke, animal intruding. Various sensors like soil moisture
sensor, fire detection sensor, ultrasonic sensors, LDR sensors etc are incorporated. The water content in the
field is detected by means of soil moisture sensor. The motor automatically switches on and off depending
on the water content and provides the information to the farmer. MQ135 sensor is used to alert the authority
in case of fire accident in the field. Rain based sensor detects the presence of rain and indicates the same,
so that the valve for water outlet can be opened. To determine whether there is enough light for the plants,
an LDR sensor is employed. Ultrasonic sensors are placed in the corners of the farm to detect any object
enters the field. The water quality is checked using a pH sensor, which alerts the user if the water is
contaminated or not. DHT-11 sensor monitors the temperature and humidity around the plant surroundings.
Deep learning is used to determine whether the plant is healthy or not. Utilizing this setup, the user can
automatically irrigate and monitor the plants which in turn increases the growth and yield of the plants.
Keywords: - IOT, Farm, Sensors, Soil Moisture, Irrigation, LDR
PAPER : EC105
DETECTING AND MITIGATING SELFISH PRIMARY USERS IN COGNITIVE RADIO
Swetha Thennavan, Thenmozhi Karuppasamy, Prem Savarinathan, Avila Jayapalan
[email protected]
ABSTRACT-- In this technology world due to the rapid growth of wireless techniques there leads to the
shortage of spectrum. Dynamic spectrum access tends to be the suitable solution to overcome the spectrum
scarcity issue and this is achieved through cognitive radio. Searching for the free spectrum or spectrum
holes is the major job of any cognitive radio. This is known as spectrum sensing. Once the spectrum hole
is available the cognitive radio can transmit the information to its intended receiver utilizing the free
spectrum. In the cognitive radio environment, there are some malicious users also known as selfish users.
This malicious user wishes to access the spectrum on its own and thus cause blockage to the cognitive users
also called as secondary users from accessing the spectrum. This attack is known as Primary User
Emulation Attack (PUEA) and it is considered as one of the important attacks in the physical layer of the
cognitive radio. Primary user emulation attack leads to inefficient dynamic spectrum access by the
cognitive radio. Hence this chapter focuses on mitigating primary user emulation attack (PUEA) effectively
through authentication tag. Threat free environment paves way to utilize the spectrum holes in an efficient
manner. The information about the spectrum holes is accepted by the cognitive radio only if it is appended
with the authentication tag. The information about the usage of the spectrum and the authentication tag is
sent to the cognitive radio by a third party who has a good relationship with cognitive radio. It is a trusty
worthy node. The trust worthy node may get some benefit or may not get any benefit from the cognitive
radio in future. It is assumed that still the trustworthy node is ready to help the cognitive radio. The
authentication tag helps the cognitive radio in making a correction decision to occupy the spectrum. The
strength of the authentication tag decides the robustness of the scheme. To enhance the robustness of the
tag generated, a double layer authentication tag generation method has been proposed. Initially Sanskrit
sutra named YRBS has been used to generate the seed value for Hash function. The output obtained using
Hash algorithm is now fed as input to Rijndael algorithm. After several rounds of processing and
142
ISBN: 979-8-3503-4821-7
permutations the double layered authentication tag is generated. By doing so the randomness of the tag has
been improved when compared to the single stage. Higher the randomness of the tag generated greater is
the time it can withstand Brute force attack. The tag so generated is combined with the spectrum free
information and transmitted by the trust worthy node. Only the intended cognitive radio can decode the
data. The other cognitive radio including the malicious node can receiver the information. But they cannot
decode the data because the key to decode the data is known only to the trustworthy node and the intended
cognitive radio. It is assumed that the key has been already exchanged between them in a secure manner.
In this way the primary user emulation attack is mitigated. NIST statistical test suite is one of the important
test suite to prove the randomness of the tag. Hence in this chapter to prove the randomness of the
authentication tag it is validated through NIST test.
Keywords: - Cognitive Radio, PUEA, helper node, Hash Function, Rijndael algorithm, Two layer
authentication tag.
PAPER : EC106
IOT BASED DETECTION OF FOOD AND WATER CONTAMINATION
R Sundar Ram, Easwar G, Thenmozhi Karuppasamy, Avila Jayapalan
[email protected]
ABSTRACT-- Food has a major role in the promotion of health and prevention of many diseases. With an
increase in the population, food quality is being decreased day by day. Nowadays, many inorganic
components are mixed with food materials, and many chemicals are used to preserve food to meet food
demand. These chemicals contaminate food and lead to many health issues. Water is one of the basic
sources for all the living organisms on the earth. Water is also getting contaminated in many ways and this
contamination results in the birth of new diseases. To overcome these problems, there is a need for a device
to detect the contamination of food and water. Internet of Things (IoT) is growing rapidly and becoming a
vast source of information. IoT has a great feature in detecting food and water quality. By connecting the
Arduino and sensors the quality of food and water can be detected and by utilizing Wi-Fi module the
information can be transmitted. Water samples are collected from various sources like tap, borewell and
wells in the local environment. Various Junk food items, fruits and milk quality is tested under food
category with the aid of sensors. If contamination has been detected then the information is passed on to
the officials in the Department of Food Safety and Drug administration for further action. In this way,
effectively checking food and water quality avoids food and water-borne illness and paves a way to lead a
healthy life.
Keywords: - IoT, Arduino UNO, pH Sensors, MQ2 gas, Contamination
PAPER : EC107
OPTICALLY RECONFIGURABLE SLOTTED WAVEGUIDE ANTENNA ARRAY FOR 5G
APPLICATIONS
Yogeshwari Panneer Selvam, Sangamithira.N.M, Thenmozhi Karuppasamy, Mihira Puvvadi, Avila
Jayapalan
[email protected]
ABSTRACT-- Rectangular waveguides with slots allows energy to escape from inside the waveguide. On
a metallic waveguide, slots of the right size and placement are used to construct a Waveguide Slot Antenna.
These slots are normally spaced out across the waveguide's narrowest face. In 5G communication, mm-
wave frequency is used to increase the transmission speed and it will also provide a large amount of
143
ISBN: 979-8-3503-4821-7
spectrum to the 5G network. A reconfigurable antenna is one that can dynamically change its antenna
characteristics in a reliable and adaptable way. Its design has been integrated with two Silicon
Semiconductor Switches to enable Optical domain Reconfiguration. Experimental results demonstrate dual
band approximately 33 GHz and 35 GHz. More numerical research revealed that the radiation pattern of
the antenna had a high gain for 5G applications. For optical network-based millimetre wave applications,
the developed optically reconfigurable waveguide slot antenna is essential
Keywords: - Waveguide Slot Antenna, millimetre wave, Reconfigurable Antenna
PAPER : EC108
Improving the Quality and performance of wireless Ad-hoc network by Using MMCN Architecture
Kalaiarasi.A, Raja.S, Kumareshan.N, Pradeepa M, Arun Vignesh, Prakash.N
[email protected]
ABSTRACT-- Mobility management and topology management plays a major role in today’s wireless
network connection. In a wireless network it consists of various types like mobile network, Ad-hoc
network, sensor network, etc. In this type of network the nodes are dynamic in nature, because of this
packet loss can be happened and it increases routing overhead. Due to this the performance of the network
get decreased, to improve the performance of the network proposed a new architecture named mobility
management using corner nodes (MMCN).In this architecture we implemented different topology like
square, pentagon and hexagonal shaped structures. In this architecture the corner nodes supports to monitor
all the mobile and it update the routing table. In experimental results compares the proposed technique with
existing technique with different corner nodes like four, five and six in an network. The proposed system
performs better based on the experimental and comparative analysis.
Keywords: - Mobility, Corner nodes, Topology
PAPER : EC109
DEEP LEARNING BASED FAKE STAMP DETECTION
Swetha Thennavan, Divya Gokul, Avila Jayapalan
[email protected]
ABSTRACT-- Philately or Stamp Collection also known as the King of Hobbies. Philatelic fakes and
forgeries are labels that look like postage stamps but have been produced to deceive or defraud. Learning
to identify these can be a challenging branch of philately. Those who produce counterfeits appeal to a very
different market from philatelists. They depend on their stamps being produced in large quantities in order
to be able to recover their outlay. Political and propaganda forgery is produced by countries in conflict to
hurt the opponent. Stamps may be issued to deprive the enemy of revenue, to distribute propaganda
material, to cause confusion, and to depict propaganda messages. Sale of fake stamp papers mean a loss of
revenue to the tune of crores of rupees to the exchequer. Stamp duty is an important source of revenue for
the government. The state governments may have to probe the large financial transactions over the past
decade to ascertain actual damage. Further, the negative effects on the financial sector would be manifold.
It is very difficult to identify a fake stamp from a real one. It takes years of experience and study to be able
to become an expert in philately and to be able to identify fake stamps. Postal services developed, early on,
measures to protect the integrity of their stamps. Some of these steps are similar to those used to protect
against forged currency. Major steps include Watermarks, Special paper, Delicate engraving, Printing
144
ISBN: 979-8-3503-4821-7
methods, Special ink for postmarks, Insertion of silk threads, Secret marks either visible or invisible to the
microscope, and Re-issue of stamps. It may not be possible to distinguish between philatelic and postal
forgery if the stamps are unused, merely by looking at them; the techniques used in producing them are
identical. Hence this paper aims on identifying whether the input given is a fake stamp. In deep learning, a
Convolutional Neural Network (CNN) is a class of Artificial Neural Network (ANN), most commonly
applied to analyse visual imagery. When an input is given, the deep learning model created will identify
whether the input is an image of a stamp or something else. If the input given is identified as image of
stamps, then it identifies whether the stamp is a real or a fake. If the stamp is real the model correctly
identifies the country to which it belongs by reading the name of the country in the stamp. Stamps belonging
to six countries over a period of 50 years were collected from the internet, and trained through the deep
learning model. As there is no dataset available for fake stamps, they were generated by modifying the real
stamps. The designed deep learning model is simple with accuracy of 99% would be very helpful in
eliminating the fake stamp.
Keywords: - Stamps, Revenue, Fake stamp, Deep learning, Convolutional neural network.
PAPER : EC110
Spectrum Sensing Framework and Energy-Efficient Resource Allocation for Cognition
Enhancement Network
Dr. Kalaiarasi Arumugam, Rajesha N, Prasad M, Shanmugasundaram N, D Sreenivasa Rao,
B.Suneela
[email protected]
ABSTRACT-- This work introduces a fire Management system to track different kinds of fire attacks
across industries. While traditional research mostly focuses on creating fire sensing. Fire alarm systems are
only useful if they can reliably and quickly send out fire alerts that include the whereabouts of the flames
the quantity of fire damage and the length of time it takes to respond to different fire alarm systems are
directly correlated. The interior environments of public buildings can be complicated and diverse, which
makes it important to consider how to safeguard people in the event of a fire and get them to a safe place
as fast as possible. Fire extinguishers , fire hydrants, , fire sprinklers, safety exit signals, smoke, fire pumps
fire doors and temperature are just a few examples of the types of firefighting equipment that may be
dynamically tracked and controlled using the Internet of Things. Recently, a few life-safety systems that
integrate a smoke sensor and an alarm device have been created for fire warning and alarm systems.
However, these fire alarm systems can make mistakes and respond to erroneous signals that aren't actually
fire signs. High-end, sophisticated fire alarm systems are necessary, and they must use a range of sensor
values to identify actual fires, including a signal from smoke detector, heat sensor, and a flame detector.
Keywords: - Fire attacks, Fire extinguisher, Fire sprinkler, fire pumps, smoke, temperature..
PAPER : EC116
Traffic Assistance and Control System- An RFID Approach
Kalaiarasi.A, Dr.L.Raja, Dr.S.Velmurugan, Dr.G.Shanthi, Mr.R.Ariharan, Dr.N.Kumareshan
[email protected]
ABSTRACT-- Websites and Web apps have become increasingly crucial over the past decade to daily life.
Over the last ten years, websites have increased from three million to over 1.7 billion. Modern centralised
145
ISBN: 979-8-3503-4821-7
digital marketplaces and enterprises provide consumers with an alternative method of selling and
purchasing items conveniently. However, disadvantages among the marketplaces include the platform’s
potential to arbitrarily block merchants, the fees associated with listing and selling products on the site, and
the lack of user data privacy. In this, we propose and demonstrate a decentralised application that takes
advantage of the Ethereum blockchain to address all these issues. The Truffle development framework was
employed in the creation of the application. An Ethereum smart contract later migrated to the Ethereum
network comprised the application’s features. The web3.js API (Application Program Interface) was used
to send the user’s input to the Ethereum network after being received through a web interface. The users
involved will be able to do transactions with the help of an interactive user interface. The interfacing is
done using the Truffle framework, and the transactions are made through MetaMask, where ETH tokens
are used. The dApp is created using JavaScript and React JS library. Thus, the integration of the blockchain
and with front-end application by interfacing it with web 3.0 helps create a secured, immutable, trusted,
and easy-to-use e-commerce website.
Keywords: - Blockchain; dAPP; ETH; web3; Truffle; Meta mask
PAPER : EC123
Development and Integration of dApp with blockchain smart contract
Dhanvardini R, Martina PA, Vijay R, Amirtharajan R, Padmapriya Pravinkumar
[email protected]
ABSTRACT-- Websites and Web apps have become increasingly crucial over the past decade to daily life.
Over the last ten years, websites have increased from three million to over 1.7 billion. Modern centralised
digital marketplaces and enterprises provide consumers with an alternative method of selling and
purchasing items conveniently. However, disadvantages among the marketplaces include the platform’s
potential to arbitrarily block merchants, the fees associated with listing and selling products on the site, and
the lack of user data privacy. In this, we propose and demonstrate a decentralised application that takes
advantage of the Ethereum blockchain to address all these issues. The Truffle development framework was
employed in the creation of the application. An Ethereum smart contract later migrated to the Ethereum
network comprised the application’s features. The web3.js API (Application Program Interface) was used
to send the user’s input to the Ethereum network after being received through a web interface. The users
involved will be able to do transactions with the help of an interactive user interface. The interfacing is
done using the Truffle framework, and the transactions are made through MetaMask, where ETH tokens
are used. The dApp is created using JavaScript and React JS library. Thus, the integration of the blockchain
and with front-end application by interfacing it with web 3.0 helps create a secured, immutable, trusted,
and easy-to-use e-commerce website.
Keywords: - Blockchain; dAPP; ETH; web3; Truffle; Meta mask
PAPER : EC124
Design development and execution of Smart Contract
Dhanvardini R, Martina PA, Vijay R, Amirtharajan R, Padmapriya Pravinkumar
[email protected]
ABSTRACT-- With the advent of decentralised cryptocurrencies such as Bitcoin and Ethereum,
blockchain technology has sparked intense interest among businesses and developers. One notable
146
ISBN: 979-8-3503-4821-7
application of blockchain is the development of smart contracts, which are code scripts that function on a
blockchain network and are automatically executed whenever the prescribed requisites are met without the
assistance of a reliable third party. As a result, smart contracts have been gaining momentum rapidly with
various applications in fields including pharmaceuticals, trade finance, supply chain management, etc. This
paper overviews the fundamental framework supporting blockchain platforms and programming
languages. This helped create smart contracts with a detailed study on creating a smart test contract running
on the Ethereum blockchain, written in Solidity programming language and executed in the Remix IDE
Keywords: - Ethereum, smart contracts, blockchain, decentralized
PAPER : EC136
Memory Effects in Digital Filters Based on Input Strict Passivity
Ronica B.I.S, Grace Kanmani Prince P
[email protected]
ABSTRACT-- The Proposed work aims to inspect and lessen the memory effects which is unusual in
filters that are digital with interference and saturation point nonlinearities through input strict passivity type
approach. New stable condition is given in order that illustrates the so-called Hankel norm execution in
filters that are digital with external interference and threshold via input strict passivity approach. The
uncommon excitations of the digital filters which happened in the past can be evaluated by the given
criterion. The new stable criterion confirms that asymptotic stability of filters that are digital on accounting
the non-availability of outward incoming signals. Thus, dominance of the discussed method is shown by a
numerical example.
Keywords: - External disturbance, Hankel norm, Linear matrix inequality, Digital filter , Saturation, Input
strict passivity
PAPER : EC149
Pest Detection and Identification on Plants Using CNN Algorithm: A survey
Kalaiarasi.A, Kumareshan.N, Kanmani.R, Dinesh Kumar.M, Dharun Pandian.R, Barath.N
[email protected]
ABSTRACT-- India is the largest agriculture production country. It ranks 2nd largest producer of rice and
wheat. Various crops are grown in various parts of the country. Agriculture contributes to India’s GDP
increase from 17.8-19.9% in 2019-2021. Larger cultivation of plants is happening on fields of India giving
many good outcomes. And we must also consider the defected plants which going to in garbage after
harvesting. So we engineers should give a solution for rectify this problem on the time of planting. Defect
in plant can identify in many views like testing the Soil nutrient, identifying the pest attack in leaves or
fruits of that plant. By analyzing the defect on the plant in cultivation time can reduce the chance of spoiling.
If the pest identification is done, then we can apply suitable fertilizer for that. Not all the insect are harmful
for plants, some will help in pollination and eat other pests on the plants. Large number of insect pest attack
leads damage in the crop production. Image processing of plant leaves for identify the pest by matching
with dataset on MATLAB Tool, eventually found the deficiency occurred on that plant. This process have
efficient algorithm for finding the result. Then the further process like pesticides applying can carried out
according to the pest identified.
Keywords: - MATLAB Tool, CNN Algorithm, Image Processing, Plant Pest Identification.
147
ISBN: 979-8-3503-4821-7
PAPER : EC151
Industry –specific Intelligent Fire Management System: A survey
Kalaiarasi.A, D. Deepak Kumar, B.Bharathraj, v.naveen Vishak, S.Jasith, L.Raja,
[email protected]
ABSTRACT-- This work introduces a fire Management system to track different kinds of fire attacks
across industries. While traditional research mostly focuses on creating fire sensing. Fire alarm systems are
only useful if they can reliably and quickly send out fire alerts that include the whereabouts of the flames
the quantity of fire damage and the length of time it takes to respond to different fire alarm systems are
directly correlated. The interior environments of public buildings can be complicated and diverse, which
makes it important to consider how to safeguard people in the event of a fire and get them to a safe place
as fast as possible. Fire extinguishers , fire hydrants, , fire sprinklers, safety exit signals, smoke, fire pumps
fire doors and temperature are just a few examples of the types of firefighting equipment that may be
dynamically tracked and controlled using the Internet of Things. Recently, a few life-safety systems that
integrate a smoke sensor and an alarm device have been created for fire warning and alarm systems.
However, these fire alarm systems can make mistakes and respond to erroneous signals that aren't actually
fire signs. High-end, sophisticated fire alarm systems are necessary, and they must use a range of sensor
values to identify actual fires, including a signal from smoke detector, heat sensor, and a flame detector.
Keywords: - Fire attacks, Fire extinguisher, Fire sprinkler, fire pumps, smoke, temperature..
PAPER : EC153
Does COVID Pandemic raise E-waste to a larger number? – Analysis andPrediction
Gayathri B, Amirtharajan R, Padmapriya Pravinkumar.
[email protected]
ABSTRACT-- The current pandemic situation, the “CORONA” era, has led to the exponential usage of
Laptops, Desktops, and Mobile phones in a larger number from March’20 to till date due to the drift in
online classes, work-from-home and online shopping scenarios. The increasing usage of electronic
components leads to a larger quantity of E-Waste disposal. According to the “Global E-Waste Monitor
Report”, 2020 by the UN, around 53.6 Mt (Million tonnes) of E-Waste have been generated globally. Asia
is number one in producing E-waste at an amount of 25Mt, and India bags the fifth position in the world.
However, E-waste collection and transportation. In this paper, initially, Regression analysis will be carried
out on the Ewaste recycling parameters like a landfill, obsolete, reused, stored and recycled and based on
the estimated value of Regression Statistics, Variance analysis and Residual Output prediction, the total
amount of E-Waste collection for the subsequent years to regulate a proper E-Waste Management System
will be predicted
Keywords: - E-Waste, Pandemic, Regression Analysis, ECMS
PAPER : EC1006
DETECTION AND CONVERSION OF CO2 INTO MOIST
Susarla S S V Suryasri, M. Satya Sai Reddy, E. Lohika Reddy, Dr.M.Naresh Kumar
[email protected]
ABSTRACT--As the atmosphere warms more and more every day due to global warming, CO2 reduction
is crucial in the current circumstances. To reduce this issue, we have come up with a solution where we
can reduce it by calculating the levels of CO2 and converting it into useful resources which will be utilized
148
ISBN: 979-8-3503-4821-7
efficiently. In this project, we have implemented the detection of CO2 levels present around us using MQ-
135 sensor using Arduino. In this we get the levels of CO2 will be shown on an LCD screen that is
connected to an Arduino board and the sensor. When the levels rise, it is a warning that we need to
efficiently recycle or convert this harmful gas. Here, the idea was to use chemicals to turn CO2 into
moisture.
Keywords: - MQ-135, LCD (Liquid- Crystal Display), CO2 (Carbon-dioxide), Arduino, LED (Light
Emitting Diode)
PAPER : EC1010
AN AUTONOMOUS CROP-CUTTING MECHANISM USING A DRONE
T.Siva Sree, Mr. Maganti Srinivas, K.Jyothi Swaroop, P.Abhishek, B.Shiva Venkata Sai Pavan
[email protected]
ABSTRACT-- various reasons are reducing the number of farmers and workers in Asia, which negatively
impacts the region's economy. It has resulted in rising labor prices, increased energy use for harvesting
inputs, and decreased resource usage. Researchers are working to find effective, low-cost, and easy-to-use
answers to these issues through the mechanization and automation of agriculture. An autonomous tractor
that can follow a specified route with harvesting equipment is built. The primary goal of this study is to
examine the autonomous tractor, which incorporates several approaches throughout its design and
construction and is navigated with drone coordinates. Here, a Hexacopter drone is used to survey the field
and relay that information to the autonomous tractor, using the data from the various sensors mounted on
the vehicle to guide it around the area
Keywords: - Autonomous Tractor, Agriculture, Drone, Hexacopter, Sensors.
PAPER:EC161
LEAN READINESS ASSESSMENT IN SUGAR PRODUCTION PLANT
Dr. P.Sundharesalingam, Dr. M.Mohanasundari, Dr. P.Vidhya Priya, M.Dhilip Kumar, R.Aarthi
ABSTRACT—The fundamental tenet of lean systems (LS) is that eliminating wastes of all kinds through
the application of lean practices. However, implementing LS without first assessing organizational
readiness could end in failure. A measurement methodology to assess the lean readiness level (LRL) has
been created in this study. The study uses the measurement framework created by Al-Najem et al., (2013)
and Nordan et al., (2010), which specify 6 constructs related to lean quality practices: process variable
(PV), planning and control (PC), customer's relations (CR), supplier's relations (SR), HR, and top
management and leadership (TML). Operational performance factors were also added to the LRL
framework. Descriptive statistics, one-way ANOVA, and regression analysis were used to create and test
1 research question and 3 hypotheses. The outcomes show that the company is far from being prepared to
apply lean. In order to succeed during complete lean implementation, the suggested LRL framework will
perform as an analytical model for evaluating quality practices before lean deployment.
Keywords-LS; LRL; CSF; PV; PC; CR; SR; HR; TML
149
ISBN: 979-8-3503-4821-7
PAPER:EC172
SURVEY ON SMART AGRICULTURE USING IOT
B.SakthiKumar,M.IshwaryaNiranjana,M.Abinath,I.GuruPrasath,K.Jugasri,S.Kailash
[email protected]
ABSTRACT—T Each day, the agriculture sector in India loses ground on both sides, which reduces the
ecosystem's potential for production. To resuscitate agriculture and return it to a path of higher growth, a
solution is increasingly needed. It takes a lot of maintenance, expertise, and management to maintain a
large-scale agricultural system. The Internet of Things (IoT) is a system of networked devices that enables
both automated job fulfilment and data transmission and reception through the internet. The crop yields are
boosted by the abundance of data analysis parameters offered by the agricultural sector. Information and
communication are modernized through the usage of IoT devices in smart farming. Soil moisture, minerals,
light, and other elements can be expected for better crop growth. This study examines a handful of these
traits for data analysis in an effort to help consumers use IoT to make better agricultural decisions in farming
and irrigation. The method is designed to assist farmers in boosting agricultural productivity.
Keywords—Smart agriculture, Irrigation, Soil, IoT, Sensor
PAPER: EC201
PERFORMANCE ANALYSIS OF STACKED NANO SHEET AT GATE LENGTH OF 10NM AT

LINEAR AND SATURATION REGIONDr. Kalaiarasi Arumugam,Prasad M, Rajesha N, Anil
Kumar V, Mohd. Abdual Muqeet, B Suneela
[email protected]
ABSTRACT: In this study, the vertically stacked Nano-sheet at the gate length of 10nm is considered. The
Nano-sheet is vertically stacked and stacked with gate oxide material of SiO2 with HfO2 each of 1nm
thickness. Three fins vertically stacked structures were examined for performance and compared same
model at lower and higher voltages. The performance comparative parameters are drive currents and short
channel effects, trans-conductance (gm), and output conductance device are considered. All simulations are
carried out using the platform of professional V-TCAD device simulator.
Keywords-Junction-less, vertically stacked fins, Ion, Ioff, Ion/Ioff Ratio, Trans-conductance and output
conductance
PAPER: EC223
IMPLEMENTING HEALTH RECORDS USING KNOWLEDGE GRAPHS AND BLOCKCHAIN
Karthika T, Mounica M, Priyadarshini N, Amirtharajan R, Revathy K, Padmapriya ,
Pravinkumar.,UmaMaheshwari B
[email protected]
ABSTRACT—Nowadays, a huge amount of patient records are being maintained, including medications,
tests, treatments, immunisations and doctor details. However, having some of these records in papers is
only possible. So, there is a growing interest towards digitalising healthcare records and maintaining
healthcare data electronically. Healthcare data maintenance is essential to create a comprehensive view of
patients and thereby improve patient engagement and health outcomes and suggests proactive measures to
counter the rising health issues. The blockchain concept is implemented for this purpose, using which the
authenticity is verified. It is done with the help of a high-level object-oriented programming language
known as solidity. Furthermore, this process creates a smart contract for healthcare data. Thus, it provides
a structure of data with inherent security qualities. In addition to the Blockchain, health records can be
visualised in a graph known as a Knowledge Graph. Knowledge Graph is written in Neo4j browser using
150
ISBN: 979-8-3503-4821-7
graph query language Cypher. This graph connects distinct quantities to represent corresponding data in a
network.
Keywords-component; Block Chain; Knowledge Graph; Patients record; Interconnected Data; Solidity;
Security.
PAPER: EC224
EDGE COMPUTATION ASSISTED GARBAGE MONITORING AND ALERTING SYSTEM
FOR A SMART CITY
Gayathri B, Akila V, Amirtharajan R, Revathy K, Padmapriya Pravinkumar., Uma Maheshwari B
[email protected]
ABSTRACT—By 2050, the majority of the world’s population will live in cities (urban areas),” predicts
the United Nations. If this prediction is true, there will be an increase in societal, environmental, and
economic challenges. To overcome these challenges and improve the citizen’s quality of life, the “Smart
City” plan is being widely adopted. To summarise, a “Smart City” model helps sustain the city and its
economy. An efficient waste management system is vital in converting a city to a “Smart City”. The waste
management system includes waste collection, transportation, and disposal. All these processes must be
monitored properly to achieve an efficient waste management system. Since the number of cloud
computing users is increasing steadily, cloud traffic is expected to quadruple in the next few years.
Additionally, traditional cloud computing technology faces serious data security threats and high
operational costs. So, storing and computing the waste collection information, status, and details of the
garbage bins in the cloud may not be a good idea. Edge computing is used to reduce the burden and
overcome the challenges of the cloud. Edge computing is also very close to the data source; this helps
reduce latency, operational cost, and bandwidth usage. Overall, edge computing provides a more reliable
system. This PAPER is used for monitoring and alerting the waste collection process using IoT (Internet
of Things), edge computing technology. The trash bins’ details and status are stored and computed in the
respective edge node. The concerned authorities can view the trash bins’ real-time status; whenever the
waste must be collected, an alert message is sent to the concerned authorities.
Keywords-component IoT; Edge computation; Smart City; Waste Management.
PAPER: EC228
SUPPLY CHAIN MANAGEMENT IN MANUFACTURING INDUSTRY USING INTERNET OF
THINGS
Chouta Jaswanth, Lanka Venkata Aravinda Krshna SaiPradeep, Charakanam Viraj Kishore,
Amirtharajan R, Padmapriya Pravinkumar.
[email protected]
ABSTRACT—Supply chain Management includes moving and storing the materials used to produce
goods, storing the finished products until they sell and tracking where sold products go so that you can use
that information to drive future sales. It includes the integrated planning and execution of processes
required to manage the movement of materials, information and financial capital in activities that broadly
include demand planning, sourcing, production, inventory management and storage, transportation and
returning excess or defective products. It relies on both business strategy, specialised software and
collaboration to work. Supply chain management is most important for any company to reduce operating
costs and maximise profits. With the help of the Internet of Things (IoT), it becomes easier to track the
151
ISBN: 979-8-3503-4821-7
state and progress of the product throughout the manufacturing process. A well-optimised Logistics
Management (Vehicle Routing) and Warehouse batching could improve production efficiency and
reduce working time, increasing profits. This PAPER aims to develop an application for Vehicle routing
with live tracking and Inventory management.
Keywords-component; IoT; Vehicle routing; Supply chain Management; Tracking; Inventory Management
PAPER: EC231
SMART SEGREGATION OF WASTE AND AUTOMATIC MONITORING SYSTEM
V. muneeswaran, Nagaraj P, A.Akhila, L.Sudeepthi , B.Venkateswararao, B.Vamsi Krishna

[email protected]
Abstract— In the present situation managing waste is becoming one of the most crucial problems across
the globe. With an intensification in urbanization, the waste generated daily in urban areas has increased.
Most of the people in rural and urban areas dump the waste in landfills directly without segregating it into
dry and wet waste. This leads to a lot of health problems and the surrounding becomes very unhygienic. If
we dump the waste directly without segregating it will increase the difficulty of recycling. To avoid all the
problems, we propose an automatic garbage waste segregation which overcomes this problem effectively.
A moisture sensor is used to segregate the dry and wet waste. The Arduino UNO controls all the activity
of sensors. Ultrasonic sensors detect the level of the object. Thus, the results have been presented by
segregating waste into dry and wet waste.
Keywords— Waste Segregation, Sensors, Dry waste, Wet waste, Arduino Uno
PAPER: EC234
IOT-BASED SMART TRASH BIN FOR REAL-TIME MONITORING AND MANAGEMENT OF
SOLID WASTE
Ayush Parth S Kumar, Basi Jithendra Sai Kumar, Bharath R L, Amirtharajan R, Padmapriya
Pravinkumar.
[email protected]
ABSTRACT—The significant growth of the Internet is leading to the emergence of new technology, the
Internet of Things (IoT). IoT is the backbone of future communication systems where everything will
communicate and share information smartly without human instructions. In smart cities, one of the
important problems is waste management. Monitoring and disposing of waste are very important to keep
the environment green and clean. This can be efficiently achieved by minimizing human involvement and
using IoT-based networks. This PAPER aims to solve the waste management issue using the Smart Waste
Mechanism (SWM). Install Smart Bins (SB) in various optimum locations and communicate with the Trash
Collecting Vehicles (TCV). Optimize their routes to the designated Smart Bin for collection whenever
prompted by the Smart Bin.
Keywords-component; IoT; SWM; TCV; Solid Waste.
PAPER: EC235
WASTE MANAGEMENT SYSTEM - A WASTAGE CLASSIFIER USING DEEPLEARNING
R.S.Sathyanarayanan, Amirtharajan R, Padmapriya Pravinkumar
[email protected]
ABSTRACT—Waste management is essential for protecting the environment, especially in the current
scenario with increasing issues such as pollution, population and land management. Thus managing this
152
ISBN: 979-8-3503-4821-7
waste is a huge task that requires a plethora of new techniques and innovation. Classification of waste into
degradable and non-degradable elements is a crucial step to be followed by everyone. However, one of the
biggest problems faced by waste management authorities is classification. The conventional method of
placing separate bins for degradable and non-degradable are not fool-proof. Segregation leads to identifying
useful degradable waste, which, when appropriately processed, can be used as manure or organic fertilisers,
which have been proven to have a positive effect on the growth of crops. This is also environmentally and
economically friendly as the waste is used appropriately instead of disposed of. Some of these are Reduced
landfill impact and Improved air and water quality. Hence a smart system needs to identify whether or not
the waste is biodegradable and then allow only that waste to be thrown into the bin.
Keywords-component; Waste Management system; classifier; Deep learning; biodegradable waste
PAPER: EC246
BREAST TUMOR DETECTION USING MACHINE LEARNING AND DEEP LEARNING
ALGORITHM
Kiruthika V, Aasthikka G, Dhivya Archana KP, Krubha Harane T, Malavika.V
ABSTRACT: Breast tumour is the most prevalent frequent malignancy inside the world, and women are
more likely to develop one. This is the world's second greatest cause of mortality for women. It accounts
for almost 22% of new cancer cases worldwide, with a 5-year survival rate of 61%. Breast cancer kills
450,000 people globally each year, according to the WHO. It is a mammography-based the most prevalent
kind of cancer in women, breast cancer, has a software detection and diagnosis method. The system is built
on the Numerous Learning paradigm, which has been used in the past by our team to create medical
decision support systems. Wavelet sub bands are used to extract the GLCM features. Then, from each
location, characteristics resulting from lesion identification (masses and micro-calcifications) as well as
textural cues are retrieved and mammography tests into "normal" and "bad" categories. The areas that
caused the automated diagnosis can be emphasised when an aberrant examination record is found. To
define this anomaly detector, two methodologies are considered.
PAPER: EC263
AUTOMATIC PRESSURE COOKER RISK DETECTION AND CONTROL USING ARDUINO

Muneeswaran V, Nagaraj P, L. Anuradha, V. Lekhana, G. Vandana, K. Gayathri Devi, A. Tejaswee
[email protected]
ABSTRACT— In this PAPER, a novel safety approach is presented to mitigate the risks associated with
pressure cookers caused by lack of understanding, ignorance, and other factors. We'll use a temperature
sensor to keep track of the temperature, which will be connected to an Arduino UNO board for control of
the gas stove, as well as a light for proximity detection and a buzzer for long distance detection. The
project's main goal is to produce a device that can be used to maintain track of the pressure cooker
temperature and automatically cut off the gas cylinder when the temperature rises.
Keywords— Arduino, DS18B20 Temperature sensor, solenoidal valve, buzzer, LED.
153
ISBN: 979-8-3503-4821-7
PAPER: EC264
SMART AUTOMATED KITCHEN SYSTEM USING INTERNET OF THINGS
Ryan Paul Jess.C, Rithin.R.S, Raguram.S , Dennis, K.MuthuLakshmi, S.Dhanasekar
[email protected]
ABSTRACT— The fundamental factor of the kitchen is security and it is one of the most prominent and
unsecured spot in every household. Day by day many household fire accidents occur and there is very little
security provided for the household people to live. Leakage of gas and fire accidents may even cause severe
injuries and even death of many people. The presence of gas leakage and fire should be immediately
identified and informed. The motivation behind this project is to make a kitchen security model which is
cost-efficient by using one of the fastest growing technology, the Internet of Things. These days the
presence of the Internet of Things (IoT) has changed the human lifestyle. IoT is a technology that makes
the objects around us to speak and 21323131313communicate. By using Arduino and the sensors we are
adding functions so that we can implement them in real time household environment to avoid accidents
like gas leakage fire.
keywords- kitchen; security; accidents; gas; fire; IoT(Internet of Things); Arduino
PAPER: EC276
A STRANGE MODEL FOR SECURE MEDICAL IMAGE TRANSACTION - AN ATTRACTOR
APPROACH
Patalam Rithish, Jagadeeshwaran Muthiah, Nithya Chidambaram
[email protected]
ABSTRACt— In the wake of the Internet, the growth of networks has steadily increased. Even though
network security is improving every day, many loopholes and gaps allow an attacker to modify, delete or
even steal data. As a countermeasure, data security is also extremely important. In the telemedicine field,
data communicated over mediums become vulnerable. Particularly Digital Imaging and Communications
in Medicine (DICOM) is the universal standard for storing and transmitting medical images, and when it
is shared among practitioners over the network must be secured. Modern cryptographical solutions are
required to address these kinds of vulnerabilities. This proposed work aims to develop a complete system
for maintaining confidentiality for medical images. The system provides robustness against various threats
such as statistical attacks and differential cryptanalysis. (Abstract)
Keywords- DICOM image, Chaotic map, Strange attractor, Image encryption
PAPER: EC278
COLD SUPPLY CHAIN ADOPTING IOT TO MAINTAIN TEMPERATURE DURING DRUG
DELIVERY
1
Vijay Ramasamy, Padmapriya Pravinkumar.
[email protected]
ABSTRACT—A pharmaceutical product, especially vaccines, consign a shelf-life predicted by
Temperature. A cold supply chain is administered with the correct Temperature from when the vaccines
are manufactured, distributed and reach the hands of public health group members and finally to the health
care providers vaccinating the public. Demand, supply and market are the key factors in PSCM (Pharma
Supply Chain Management). Shelf life prediction of drugs adopting TTT (Time Temperature Tolerance)
helps maintain the key factors in PSCM. This PAPER focuses on maintaining the correct Temperature
during drug delivery by utilizing IoT (Internet of Things). A case study is considered and implemented in
the Sun Pharmaceutical industry in maintaining the shelf life of the vaccines like Lupride and Degapride
in considering the chemical reactions and microbial growth.
Keywords-component; Cold supply chain, Temperature, Pharma, PSCM, Shelf life, Vaccines, IoT
154
ISBN: 979-8-3503-4821-7
PAPER: EC279
REINFORCEMENT LEARNING ALGORITHM TO SUSTAIN TEMPERATURE IN PHARMA
SUPPLY CHAIN MANAGEMENT
Vijay Ramasamy, Padmapriya Pravinkumar
[email protected]
ABSTRACT—The cold supply chain in the pharmaceutical industry is a technique designed to transport
and store temperature-sensitive vaccines to sustain their effectiveness and to uphold safety measures as
they are directly involved with human life. The reinforcement learning algorithm is a feedback mechanism
using machine learning where the agent, environment, action, state and rewards are correlated. This PAPER
is to sustain a cold supply chain for vaccines like Lupride and Degapride in Sun pharmaceuticals; a
reinforcement learning algorithm has been devised to measure, monitor and correct the temperature of the
cold storage unit. Temperature-sensitive vaccines need to maintain their chemical and biological traits
during their transit from packaging to delivery.
Keywords-component; Reinforcement learning, Temperature, cold supply chain, Pharmaceuticals
PAPER: EC287
Electronic Voting Machine Using Fingerprint Scanner with Image Enhancement Technique
Muthukumar ArunachalaM, Sai Deepak AlapatI, Chandana Chennnamsetty3 Pujitha Dantam4 Anusha
Dabbara5
[email protected]
ABSTRACT — Elections are very important to decide country’s future. Manual voting system have many
consequences, to overcome these problems implementing biometric voting machine in society diminish
problems. Hence, a fingerprint-based voting system has been proposed in this paper. Every person has
unparalleled identity by using this bit new system works. In this proposed system, fingerprints are collected
during enrollment phase, and it will be verified during polling. For this purpose, the hardware component
Arduino Uno has utilized in this work. The Adafruit library tells the sensor to read fingerprints, store them
as “templates”. The voter has to place his finger to register, that data will be stored in software. Persons
who had registered or already existed in list can have access to cast their vote by placing their fingerprint.
If any person tries to fake their vote by casting second time or non-registered the buzzer rings. By this
change voting system get rid of some issues.
Keywords — Electronic Voting, R305 Fingerprint Module, 1602 Parallel LCD Display, Arduino,
ATmega328
PAPER: EC297
DESIGN AND DEVELOPMENT OF UAV FOR CHALLENGE FACING FOR DATA MULING
PROBLEM IN PATH
Muthukumar Arunachalam, Mani Kanta Animisetty, Bhanu Prakash Nammi, Suman Jajala, Siva
Raju Dokku
[email protected]
ABSTRACT — In order to evaluate the utility of using UAVs for the ongoing collection of sensor readings
from sensor nodes situated throughout an environment and their delivery to base stations for additional
processing, this PAPER addresses the problem of traffic engineering in Internet of Things (IoT) settings.
UAV allocation and persistent path planning model has been proposed in this PAPER, where data from
ground sensors is gathered by a group of heterogeneous UAVs from different base stations, and the
information is then sent to the base stations that are nearest to the sensors. This issue has been
mathematically shown to be nondeterministic polynomial hard as a model of constrained real-time
155
ISBN: 979-8-3503-4821-7
optimization. In this research, a heuristic solution to the issue is put forth, and its relative effectiveness is
assessed by experiments on both synthetic and actual sensor networks, utilising various UAV settings.
PAPER: EC298
LP/CP RE-CONFIGURABLE ANTENNA FOR 5G APPLICATIONS
Rangarao Orugu,M. Nesasudha,Doondi Kumar Janapala
ABSTRACT— this paper, presents a polarization re-configurable antenna for 5G applications operating
at 3.5GHz for n78 band. A micro-strip fed rectangular patch antenna is used in this design. A slant
rectangular slot is made into the patch to divide it into two parts. One part is connected to the strip feed
and another part with gap made due to slot. The non fed part of the patch is connected to an E-shaped
stub in a slant way similar to the slot. A PIN diode is placed into the slot to make connection in between
the two parts. The PIN diode switching condition helped in shifting polarization from linear to circular
vice versa, making the antenna suitable for 5G applications.
Keywords- circular polarization, linear polarization, re-configurable, PIN diode, 5G antenna
PAPER: EC299
CUSTOM AND DESIGN OF AGRI-DRONE
Dr.A.Muthukumar, Muthukumar M V, Tamil Varshini S,Prem Mathavan N, Vishnu K
[email protected]
ABSTRACT--The primary reason for using the proposed Agricultural drone is to spray fertilizer. Making
use of 3D printing technology, the Agri drone is created. Poly lactic acid (PLA) is used in the drone's body
construction. The electronic components used for drone are flight control (KK2.1.5), electronic speed
controllers (ESCs), transmitter and receiver (six channels), BLDC motors (1000 KV), propellers
(1045),relay and a 12v dc pump. Electronics parts and 3D printed pieces are used to assemble the
agricultural drone. It is urgently necessary to update old agricultural practices due to the world's population
growth and shrinking arable land. The problem of effective and productive farming is serious on many
levels, particularly in a country like India, which was once regarded as the "land of farmers." The only
answer to this issue is modernization of agricultural methods. Innovations in biotechnology, appropriate
instruction, and the use of cutting-edge farming equipment can all help to achieve this. Drones for
agriculture might be one such instrument. We've described a Do-It-Yourself (DIY) method for creating
your own agricultural drone in this study. Assembly mistakes, sizing problems, and weight balancing
problems were challenges faced throughout the drone's development. To create the agricultural drone,
modern technologies including 3D printing, rapid prototyping, and drone technology are applied. An open
environment is used to test the Agri drone's ability to spray fertilizer.
Keywords—unmanned aerial vehicle;KK2.1.5;3D printed drone; drone spraying; PLA-based drone;

aerial vehicle, agricultural drone
PAPER: EC300
DUAL SEMI CIRCULAR DISC LOADED MONOPOLE ANTENNA FOR 5G/WLAN
APPLICATIONS
Rangarao Orugu,M. Nesasudha,Doondi Kumar Janapala
Abstract—This paper presents a monopole antenna operating at 3.6GHz for 5G applications. The design
consists of a planar monopole strip line connected to two semicircular discs on both side of the strip line.
The FR4 substrate is used as dielectric material and the atrial ground is used at back side. The ground plane
is maintained in a Ω shaped conducting part. The antenna operates at 3.6GHz with bandwidth of 110MHz.
156
ISBN: 979-8-3503-4821-7
The antenna maintained a compact size of 28mm×20.5mm. The antenna is suitable for narrow band 5G
application at 3.6GHz.
Keywords- Monopole antenna, 5G, 3.6GHz, circular discs
PAPER: EC309
MODIFIED ENCRYPTION STANDARD FOR REVERSIBLE DATA HIDING USING AES AND
LSB STEGANOGRAPHY
B Elisha Raju, M Ravi Sankar, T Saran Kumar, K Ramesh Chandra, V Bhavani Durga , G Prasanna
Kumar
[email protected]
ABSTRACT—With the advancement of cloud technologies and data privacy protection, reversible data
hiding in encrypted images has gained more attention, as technology can embed additional data in the
domain of encryption, ensure that the embedded data can be extracted error-free, and restore the original
image without losing any information. The main aim of this project is to conceal the data in the image with
schemes of combing reversible, encryption mechanism and Auxiliary syndrome. The original image is first
encrypted using an encryption key, after which the auxiliary syndrome for the first half of the original
image is generated. The message is inserted into the first sub-image of the cipher picture after image
encoding and encryption, and each pixel can load a bit of secret message. The embedded data can be
effectively retrieved and the original picture can be correctly restored on the receiver side utilizing the
encryption key and extracted auxiliary syndrome. The proposed method gave the best analysis in terms of
Peak Signal to Noise Ratio (PSNR) and Bit Error Rate (BER).
Keywords-cryptography, cipher-image, auxiliary syndrome, segmentation, lsb modification, pixel;
PAPER: EC323
A FLEXIBLE WIDEBAND SPIRAL ANTENNA USING POLYDIMETHYLSILOXANE

SUBSTRATE FOR WEARABLE APPLICATIONS
V. A. SANKAR PONNAPALLI, GAYAM SHIVAKANTH REDDY, PEDDI SREE ROHITH GOUD, PABBA SAHARSH,
PRAVEENA A
India.*[email protected]
ABSTRACT— This work presents a flexible wideband antenna made of transparent
polydimethylsiloxane (PDMS) substrate. The proposed antenna’s overall size is 28x24x1.5 mm3. The
proposed wideband antenna operates from 3.25GHz to 15GHz and has good gain and an omnidirectional
radiation pattern. The Specific Absorption Rate (SAR) values are also calculated when it is fed with a
power input of 50mW and the values are found to be below the recommended FCC limit. It has been
examined at a few different frequencies 3.6GHz, 4.2GHz, 4.8GHz, 5.2GHz, 5.8 GHz, 6.2 GHz, 7.5 GHz,
8.4 GHz, 9.2 GHz, and 10GHz. The maximum achieved gain is 6.56 dB at 3.6 GHz. The performance
characteristics of the proposed antenna without bending, while bending and when placed on human tissue
are compared.
Keywords- Polydimethylsiloxane (PDMS), Wearable, Flexible, Wideband, Specific absorption rate (SAR),
FCC limit.
157
ISBN: 979-8-3503-4821-7
PAPER: EC324
IOT-Based Smart Inventory Management System
Kiruthika J K, Manikandan, Nithya U, Preethika M, Poorna S
[email protected]
ABSTRACT--With the rapid advancement of technology, tasks in our lives are becoming increasingly
simple. The most important aspect of efficient retail management is inventory control. The world is moving
toward a more intelligent future in which everything is automated and remotely controllable by using sensor
technologies. This process not only takes time but sometimes results in mistakes that were not intended.
Large storerooms present a particularly complex situation that makes it exceedingly challenging for
humans to manually check the inventory on a regular basis. The well-known NodeMCU ESP8266
microcontroller, which is based on the Internet of Things and sensors, can properly provide the weight of
a specific good as well as supports high in automatically counting the number of items on a shelf.
Keywords: Internet of things, Technology, Inventory.
PAPER: EC334
DESIGN AND ANALYSIS OF ANNULAR SLOT ANTENNA WITH MONOPOLE FOR WLAN
USING COMPUTER SIMULATION TECHNOLOGY
Thakur Rithika, Pathuri Sree Mahalakshmi, Brahmandlapally Anil, N. Arun Vignesh, Dr.
Balambigai Subramanian
[email protected]
ABSTRACT—Antennas are fundamental for wireless communications. Portability, affordability,
adaptability, and small size are essential for wireless communication antennas. This PAPER discusses the
designing and analysis of an Annular Slot Antenna (ASA) with a monopole for Wireless LAN (WLAN)
applications at 5.8GHz frequency using Computer Simulation Technology (CST). The suggested antenna
has a basic construction and is simple to build. A monopole and an annular slot antenna (ASA) are fitted
into each other to form two quasi-omnidirectional radiators. A microstrip line supplies power to the
monopole and the ASA. The performance of the designed antenna is analyzed on the grounds of voltage
standing wave ratio (VSWR), S parameters, Return loss, and radiation patterns. Simulated outcomes from
CST are presented.
Keywords—Annular slot antenna, Computer simulation technology, a monopole antenna, WLAN, VSWR
PAPER: EC335
DESIGN OF DUAL BAND BODY AREA NETWORK ANTENNA USING CST
Randhi Harika, Manda Arun Raj, Saraswatula Karthikeya Anirudh, N. Arun Vignesh, C. Gokul
Prasad
[email protected]
ABSTRACT-In this paper a wearable antenna is proposed for WBAN frequency range. The antenna is a
dual band button with a diameter 14.52 mm integrating a patch on top of a dielectric section. The button is
located on top of a fabric substrate and a conductive fabric ground that are to be incorporated in wearables.
158
ISBN: 979-8-3503-4821-7
Computer simulation technology was used to create the entire topology. This antenna's key peculiarity is
that it emits two different radiation patterns: a monopole-type and broadside pattern for 2GHz and 5GHz
band respectively. A scaled-down prototype has been created. The results and consequences of real-world
applications are observed, quantified, and remarkably consistent with models.
Keywords— button antenna, wearable antenna, textile antenna, cst, dual-band, on/off-body radiation
pattern
PAPER: EC337
DESIGN AND ANALYSIS OF M-SHAPED MONOPOLE ANTENNA FOR BLUETOOTH
COMMUNICATIONS
Andrews Jogu, Kakinada Leelavathi, Iruvuru Manogna, N Arun Vignesh, Dr. Balambigai
Subramanian
[email protected]
ABSTRACT—The M-shaped monopole antenna is an antenna that has a reduced size for bluetooth and
wireless communication. It is an integrated, portable, wideband antenna with a gain of 2dBi–5dBi and a
frequency range of 0.8GHZ–2.45GHZ. A 0.4mm substrate height and 4.4 relative dielectric constant was
found to have a VSWR< 3:1. The T-shaped groundplane is incorporated into the design at the price of
bandwidth in order to further minimize size, which was chosen to represent the dimensions of a normal
mobile phone. Two trapezoidal sections, a triangular part, and a rectangular section are combined to form
the M shape of the M-shaped antenna. A 50Ω feed line can be used as the impedance bandwidth is huge.
Over the operational band, the radiation pattern is comparatively steady, producing an omnidirectional
pattern. The antenna is a desired component for usage in a variety of mobile communication systems due
to its small size and broad bandwidth.
Keywords- M-Shape, Wideband, Bluetooth, Communication, Integrated, Wireless.
PAPER: EC338
STEGANALYSIS USING CONVOLUTIONAL NEURAL NETWORKS YEDROUDJ NET
Anuragh Vijjapu, Y Satya Vinod,S V S N Murty, B Elisha Raju, B V V Satyanarayana, G Prasanna
Kumar
[email protected]
ABSTRACT—Since vestige, steganography has existed in securing delicate data against unapproved
sensational activities. By and by, advanced media's development, uncovers that steganography has been
utilized as a device for exercises like psychological oppression. Considering this foundation, steganalysis
emerges as a cure to steganography. Steganalysis can be separated into two principle draws near: all
inclusive - likewise called blind - and explicit. Explicit strategies demand a past information on the
steganographic strategy under investigation. Then again, widespread techniques which can be broadly
rehearsed in an assortment of calculations are more versatile to certifiable applications. Hence, it is
important to build up much more exact steganalysis strategies equipped for recognizing the secret data
coming from the utilization of different steganographic techniques. Taking into account this, a general
steganalysis technique i.e. yedroudj network worked in pictures is proposed. The proposed methodology
depends upon the common counterattack process of steganography in which the data extraction and the
classification process involves. The performance of the proposed methodology is carried on the stego
images generated by the tools like xiao and openpuff. It just so happens, the proposed strategy prevails
generally, giving good outcomes on shading pictures and promising outcomes on dim scale pictures.
Keywords-steganography; yedroudj; openpuff; xiao; data hiding
159
ISBN: 979-8-3503-4821-7
PAPER: EC352
IMPACT OF MOBILITY AWARE RPL OBJECTIVE FUNCTIONS IN MOBILE INTERNET OF
THINGS APPLICATIONS
V. C. Diniesh, G. Murugesan, M. Joseph Auxilius Jude, S. Anantha Kumar, K. S. Dharani, R.
Aravind
[email protected]
ABSTRACT— Data routing is one of the most important difficulties in these networks because of the rise
of IoT applications, which results in the processing of a vast volume of data. A major factor in IoT networks
is low-power and lossy networks, hence RPL was created to address these issues. Mostly IoTs applications
involves in static for past decade and to meet with future applications mobility and dynamic topology
changes makes more challenging to researcher need to progress the disconnection of nodes and minimize
loss. This PAPER we investigate the congestion control and load balancing of RPL routing under different
scenarios and dynamic traffic condition. Finally the performance evaluation is carried out in existing RPL
objective function (OF) and to validate its effectiveness the routing metrics are considered in terms of PDR,
delay and packet overhead.
KEYWORDS: RPL, ROUTING, WSN, IOT, MOBILE ROUTING.
PAPER: EC361
BLOCK LEVEL IMPLEMENTATION OF ORCA TOP USING H-TECHNIQUE
Gattu Harish Goud, Dr.T.Padma , Dr.Ch.Usha KumarI, Dr. Kalaiarasi A
[email protected]
ABSTRACT—An implementation to design Block level clock tree of a complex system-on-chip

practically using ORCA Top technique. A highly integrated but low power System-On-Chip with fast
design productivity and low development cost is main motto of this project. Here it was introduced a
flexible Block Level clock network topology and also its synthesis algorithm used a H-Technique to search
for near optimum solution in shorter turn-around time. In the present work designed with 28nm System-
on-Chip product which shown that the topology and algorithm managed to produce an averagely 16.98%
better global skew, 42.75% less divergence on critical clock paths and with 64.5% shorter clock balancing
phase compared to a conventional ASIC methodology used earlier.
Keywords— Floor plan, Power plan, Placement, Clock Tree Synthesis, Routing, Slack, Design Rule
Check.
PAPER: EC371
IMPROVED RIEMANN SEQUENCE BASED SELECTIVE MAPPING
Chhavi Sharma, Pankaj Sharma
[email protected]
ABSTRACT--Orthogonal frequency division multiplexing (OFDM) is recently introduced as a waveform
candidate for6TH generation New Radio (5GNR) system and it is also popular for5th generation wireless
communication systems due to various advantages. Large peak to average power ratio is a major drawback
of OFDM signal. It is necessary to reduce PAPR of OFDM system for power efficient operation of high
power amplifier. This PAPER presents an improved Riemann sequence based SLM (IRSLM) for PAPR
reduction of OFDM system. From the results, it is revealed that the PAPR of proposed scheme is improved
by 1.9dB, 0.9dB and 0.4dB compared with CSLM, WSLM and Riemann based SLM (RSLM), respectively.
Keywords: PSD, GFDM, OFDM, Polynomial
160
ISBN: 979-8-3503-4821-7
PAPER: EC378
IMPLEMENTATION OF BEAM SELECTION ALGORITHM IN MILLIMETER WAVE MIMO
SYSTEMS
CH V V S Srinivas, N Durga Naga Lakshmi, A K Chaitanya Varma
[email protected]
ABSTRACT— Beam selection is an important topic in modern wireless communications. In a millimeter

wave system, a traditional beam selection idea is used to reduce the number of required Radio Frequency
(RF) chains. Consider the downlink multi-user multiple-input multiple-output (MIMO) systems at the
access point, which operate at millimeter frequencies (30 GHz - 300 GHz) and are equipped with a high
number of antenna components. Beam space MIMO with beam selection is an appealing technique for
lowering the number of radio frequency chains necessary in such systems in order to minimize energy
consumption while providing higher data rates and improved signal coverage.
The proposed algorithm is a heuristic greedy algorithm that employs low computational complexity with
moderate performance. By simulating beam selection as maximum weight matching on a bipartite graph,
and also beam selection based on the Kuhn-Munkres algorithm, introduce a new formal framework. These
two proposed algorithms offer an efficient way of selecting beams, especially in sparse systems with low
power consumption.
Keywords- Beam selection, Radio frequency, Millimeter wave, Multiple input multiple output, Sparse
system.
PAPER: EC444
HIGH PRECISION NON-INVASIVE BLOOD GLUCOSE MONITORING DEVICE FOR
DIABETIC PATIENTS
Roshini.Y, P. Durgaprasadarao, Sai Sarvan. R, L. V. Koushik Varma U, V.Sireesha
[email protected]
ABSTRACT --One of the leading causes of sudden illness and mortality among non-contagious diseases
is diabetes. In their daily lives, people with diabetes need to strike a balance between the three crucial
components of food, activity, and medicine. As a result, it is essential to continuously check blood glucose
levels when treating diabetes. The state of the diabetic patient can be improved by lifestyle changes that
control blood sugar, cholesterol, and blood pressure. Using a self-monitoring glucose meter is the preferred
way for determining blood glucose levels. which is intrusive, uncomfortable, and expensive. As a result,
there is an increasing need for non-invasive, dependable glucose monitoring methods. The development of
non-invasive blood glucose monitoring technologies has attracted the attention of researchers from all over
the world.
Keywords: Non-invasive, Max30102, Blood Glucose, Transmittance, Reflection.
PAPER: EC480
METAVERSE: CRYPTOCURRENCY PRICE ANALYSIS USING MONTE CARLO
SIMULATION
Divya, Anisha Singhal, Neha Singhal
[email protected]
ABSTRACT—Among the most recent and quickly growing industries worldwide is metaverse
cryptocurrency.Despite the fact that the first cryptocurrency namely Bitcoin was only developed 13 years
ago, the use and value of digital currencies have increased dramatically. Their marketplaces are among the
most intricate and unpredictable markets over the globe. Despite being a decentralised system that does not
161
ISBN: 979-8-3503-4821-7
have a single point of failure, it is commonly known that many individuals who invest in it lose their money.
In this article, we attempt to use Monte Carlo simulation to forecast the price of a selected few well-known
cryptocurrencies for the next 1000 days namely Doge coin (DOGE), Ethereum (ETH), Litecoin (LTC) and
HEX.The mean return on a cryptocurrency and the standard deviation of past returns are the two key factors
that affect how the simulations end out. Additionally, the study evaluates the projected returns for
cryptocurrencies as well as the correlation between all the different currencies. It also looks into the
statistics of cryptocurrency, average true range, volatility and cumulative returns. By calculating the growth
rate and sharpe ratio, the entire research would also assist investors in deciding if they should purchase a
specific cryptocurrency or not.
Keywords—Metaverse, Cryptocurrency, Monte Carlo Simulation, Volatility, Investment returns, Price

Prediction
PAPER: EC491
ANALYSIS OF MEMRISTOR BASED ANALOG CIRCUITS FOR IMPROVED EFFICIENCY
Prabhnoor Bawa, Vandana Niranjan
[email protected]
ABSTRACT—Memristor is the fourth fundamental electrical component that bridges the missing link
between charge and flux. It is a nonlinear, passive device that has garnered attention in various areas of
circuit design, because of its unique memory storage capabilities in the absence of a power source. In this
paper, analog circuits of an adder, subtractor, and active low-pass, band- pass, and high-pass filters, have
been modelled using memristors. Compared to the resistor-based circuit, the power dissipation from the
memristor-based model was reduced by 9.52% for the arithmetic circuit and by 6.68% for the active filter.
All circuits have been implemented and simulated on LtSpiceXVII.
Keywords-memristor; memory effect; adder; subtractor; active filters; SPICE Model; power consumption;
low-pass filter; high-pass filter; band-pass filter.
PAPER: EC496
A TRANSFER LEARNING APPROACH TOWARDS OPTICAL SPECTRUM BASED OPTICAL

PERFORMANCE MONITORING
Sindhumitha Kulandaivel , Rithik Yadav, Rahul Anand, R. K. Jeyachitra,
[email protected]
ABSTRACT—In this paper an attempt is made to perform the comparative study of transfer learning
models for optical performance monitoring. The optical spectrum of 64 QAM under varying filter effects
are collected as the input image dataset which is then boosted with data augmentation to increase the
training data. Then the transfer learning models are used for monitoring of Optical Signal to Noise Ratio,
off-center filtering effect and the cascade filtering effect. The results obtained are compared with a newly
built CNN model to compare the advantage of utilizing TL. The obtained results indicate that the TL models
yield better accuracies in shorter period of time using a smaller training data with the VGG-19 model
performing well with an accuracy of 94.7% for the proposed application.
Keywords—transfer learning, optical spectrum, modulation format recognition, off-center filtering effect,
cascade filtering effect
162
ISBN: 979-8-3503-4821-7
PAPER: EC497
A SMART AUTO IRRIGATION SYSTEM – AN IOT BASED APPROACH

Ramya Vijay, Yedupati Vishnu Priya, Praveen Chandra Reddy, Monisha S, Vijay Ramasamy
[email protected]
ABSTRACT— Internet of Things (IoT) plays a crowning role in smart agriculture. Smart irrigation
provides precise and economic irrigation with lowest manpower. This project will help root level farmers
to get into smart irrigation system. Smart irrigation is an empirical concept employs IoT sensors capable of
providing information about their agriculture fields and making irrigation automated by Internet of Things.
The significance of the proposed solution includes soil characteristics, atmosphere characteristics and the
crop type to schedule the irrigation. In addition, the proposed system confirms the irrigation accounting
weather forecast dat. A decision tree prediction algorithm is employed to analyze the sensed data to decide
the irrigation. A microcontroller-based controller unit communicates with web server thru wireless
communication and turns ON/OFF the motor based on the decision.
Keywords- Internet of Thing; Smart Agriculture; Agri-Tech; Irrigation; sensors
PAPER: EC500
RELATING DT FRAMEWORK IN IOT BASED HEART RATE AND BLOOD OXYGEN
MONITOR WITH AUTOMATIC DATA SAVING
Gokul Prasad C1, Vipshal W G2, Leon O3, Thamarai Kannan AP4, Sabarinathan B5.Arun Vignesh
N6
ABSTRACT—It might be challenging to monitor and keep tabs on a patient's health when they are at
home. Patient’s health state should be routinely assessed and reported to the relevant doctor so that
any necessary emergency action can be taken, especially in the case of elderly patients. In this paper,
we suggest a system that, monitors patient health with sensors and notifies a patient's doctor or family
members via the internet in case of an emergency. The strategy was created for people who don't
require emergency care but still require routine supervision from a physician or a family member.
Utilizing a temperature and heartbeat sensor, our system keeps track of patient health. It primarily
focuses on circumstances in which patients and doctors are separated by great distances and in which
it is critical to provide all relevant patient information to the physician. The system promptly alerts the
doctor about the patient's status over an IOT device whenever it detects any abrupt changes, such as
when the body temperature exceeds the set limit or when the heartbeat is irregular. The doctor can take
the appropriate action based on the data received by the smart patient health monitoring system. The
information that has been given to the doctor can be protected from outside interference by using block
chain technology and smart contracts. This was incorporated and implemented with Design Thinking
(DT).
Keywords— IOT, sensors, block chain, and health monitoring Design Thinking.
PAPER: EC503
A DT Framework based AI Robot For Facial Recognition using IoT
Gokul Prasad C, Gowtham S, Mahesh M, Azeem Biju Sainudeen, Prabhakaran P, Kumareshan
[email protected]
Abstract: One of the main objectives in the modern technological world is human facial expression
recognition. Almost all fields are using robotic applications. Emotion recognition is essential for enabling
163
ISBN: 979-8-3503-4821-7
successful human-robot interaction. The goal of this project is to create and put into use a revolutionary,
automatic facial recognition and AI (Artificial Intelligence) system for emotion detection (Internet of
Things) incorporating Design Thinking.
Keywords: Face expression, Emotion Recognition, Artificial Intelligence, Internet of Things, Design
Thinking
PAPER: EC509
Performance Analysis of Low-Density Parity Check Codes for 5G Technology
Pooja Pathak, Richa Bhatia
Abstract: In the contemporary world, the need for mobile internet has been increasing progressively, and
this increase in demand forces an urgent requirement for 5G wireless communication. In comparison to
4G, the modern 5G cellular system is described by three sets of key use cases, which are Enhanced Mobile
Broadband (EMBB), Ultra-Reliable and Low Latency (URLLC), and Massive Machine Type
Communications (MMTC), and all these cases require higher throughput, reduced latency, and high
reliability. Channel coding improves the efficiency of wireless communication by adding some extra bits
to the information bits to be transmitted in a controlled manner, thereby increasing the efficiency of the
communication. The two-channel codes that are receiving more and more attention because of their
capacity-achieving capability are Low-Density Parity Check (LDPC) and Polar codes. LDPC codes have
been chosen as the candidate codes for data channels because of their high parallelism encoding and
decoding, which can realize higher throughput encoders and decoders, whereas polar codes have been
selected for control channels as control signals are transmitted at a lower rate and these codes offer good
performance in the low block error rate regime.
Keywords: 5G, Channel codes, LDPC codes, Bit error rate, Parity check matrix.
PAPER: EC539
Identification of Gender and Age using Classification and Convolutional Networks
D. Durga Prasad, K.V. Subba Rami Reddy, N.Akash, K.Dakshayani, N.Jessica, Ch. Rama Krishna
[email protected]
Abstract --- Today's technology advances rapidly. A person's face is extremely important in helping to
recognise them specifically. When seen from a human viewpoint, it is fairly simple to determine a person's
gender and age just on visuals, but when viewed from a computer perspective, it is not possible. The main
goal of this project is to create an algorithm that accurately predicts a person's age and gender.HAAR
cascade is one of the most used methods. Using HAAR Cascade, create a model for this project that can
determine a person's gender. Different male and female photos that were both positive and negative were
used to train the model's classifier.Various face traits are taken off. The input image's gender will be
determined by the HAAR Cascade classifier with its help. Deep Convolution neural network is used in it.
It functions well even with little data.Use the Caffe deep learning framework for the age estimation
challenge. Caffe offers extendable code and expressive architecture. Caffe can handle more than 60M
images per day. It is one of the quickest convent implementations available as a result.
Keywords-- Gender recognition, Age classification, HAAR cascade, Caffe deep learning framework,
Neural network, Convolutional Neural Network.
164
ISBN: 979-8-3503-4821-7
PAPER: EC541
Performance Analysis and comparison of Improved MFCC with HMM/VQ modeling for speech
recognition
Kalaiarasi.A, Dr.A.Narendrakumar, P.Jayabalasubramaniam, Kumareshan.N, Prakash.N, Gokul
Prasad.C
[email protected]
Abstract—This paper focuses on performance analysis and comparison of different extraction methods
like Bark frequency cepstral coefficients (BFCC) Mel-frequency cepstral coefficients (MFCC) and Linear
predictive coefficient cepstrum (LPCC) for speech recognition system. HMM/VQ classification method is
used for different feature extraction methods as it performs better than other methods. Comparison is done
based on average recognition accuracy in noisy environment.
Keywords— Feature Extraction Methods,Classification method, Speech Recognition, HMM/VQ.
PAPER: EC566
Smart MCB Design using Embedded and IOT system
Ayush Yadav , Vijay Kumar Sharma
[email protected]
Abstract: This paper gives a unique method for the replacement of conventional miniature circuit breaker
using embedded system and internet of things. The model which has been described in the paper will not
only replace the conventional MCB but also provides much better functionality as compare to conventional
one. The automation tool used in this model (a programmable software) connects the email account with
this device so that user can get notification in their email account when the voltage become extremely high
(above 240 V, the value of extremely high voltage may vary from user to user and it can be changed as per
the user) and they will continuously get mail notification until and unless the voltage drops to below the
set level. Not only the user will get notification but all the electrical appliances which are connected with
this device will automatically turn off when voltage becomes extremely high. When the voltage supply
level starts to decrease then those appliances which are getting the voltage within their desired range will
automatically “ON” without user interface. By using “N” channels relay with the model, “N” number of
electrical appliances can be connected with this device. The device can operate the appliance with different
voltage requirement at the same time.
Keywords: Programmable microcontroller, electromagnetic switch, IFTTT, Email, AC Voltage Sensor,
IOT System, Arduino IDE Software, Electrical Appliances
PAPER: EC574
Deep Learning Framework for Identification of Leaf Diseases in Native Plants of Tamil Nadu
Geographical Region
Dr. K. Kavitha, S. Naveena
[email protected]
Abstract—Plant pathogens are a prominent cause of reduced yields, resulting in decreased crop yields.
Scientists are striving to develop a mechanism for identifying plant ailments in order to boost farm output.
Deep learning algorithms have been developed for pathogen recognition and prediction in tomato plant
leaves. Two different types of diseases impact both healthy and sick leaves. A Convolution Neural
165
ISBN: 979-8-3503-4821-7
Network, which is effective for detection and prediction barrier, was used to forecast Septoria spot and
bacterial spot. A dataset of 4930 images of healthy and damaged leaves from a plant community is used
for the experiments. The model's performance is precisely evaluated, and the conclusion is accurate. The
project makes use of Plant Village images of tomato, potato, and onion leaves. Four different classes can
each be recognized by the suggested CNNs. In each instance, the trained model achieves accuracy of 100%,
98.3%, and 97.89%. The classification of leaf disease detection using simulation data shows the potential
effectiveness of the proposed approach. The algorithm proposed can be applied to categories any additional
species of native plant to Tamil Nadu. Self Help Groups (SHGs), which are found in each and every village
in India, will be utilized to gather information on how farmers see themselves. The observations and
ameliorate both will be communicated to the same SHGs. Because of its high success rate, the model is a
good tool for counselling or early warning.
Keywords—Image processing, Convolution Neural Network, self Help Group, plant leaves, plant diseases.
PAPER: EC581
Design of IoT and Machine Learning based Model for Crop Prediction and Fruit Ripeness Detection
G. Saranya, N. Dineshkumar, A. S. Hariprasath, G. Jeevanantham
[email protected]
Abstract— Being an agricultural country, India's economy is heavily dependent on the growth of
agricultural products and related market goods. In India, rain, which is incredibly unpredictable, has a major
impact on agriculture. The development of farming also depends on a variety of requirements, especially
nitrogen, phosphorus, potassium, plant growth, moisture, surface temperature, and meteorological
conditions such as temperature, rain, and so on. India is now making rapid technical progress. Innovation
will therefore be beneficial to agriculture, improving plant efficiency and providing farmers with
significantly higher yields. The proposed employment provides a service for wise farming by
recommending plants to the agricultural sector, which can greatly increase farmers' productivity. Weather
report information acquired from IMD (Indian Metrological Division) such as temperature level and rains
and specifications database provides an understanding of which crops are appropriate to be cultivated. This
work provides a system, for the development of an android-centered application, which utilizes information
analytics methods in purchase to anticipate one of the most lucrative plants in the present weather condition
problems. The suggested system will incorporate the information acquired from the database, and weather
condition division, and by using the artificial intelligence formula: Synthetic neural network, a forecast of
many appropriate crops inning accordances with present ecological problems is made. Everybody is
worried about their health and wellness and is pocket money to remain in healthy. There is a range of food
products offered that make one healthy and balanced and fruits are among them. The usage of fruits is so
essential that it includes great deals of minerals and vitamins that are needed for an individual to remain
healthy. However, nowadays these fruits are being ripened utilizing synthetic ripening representatives
which are hazardous to our health and wellness and can result in different illnesses. This paper provides an
affordable technique that can acknowledge whether a climacteric fruit is ripened artificially or otherwise.
Deep Discovering methods are utilized to determine whether the fruit is ripened artificially or normally.
Along with this fruit, ripening is due to natural ripening, or chemical is also detected using machine
learning.
Keywords—IoT, Machine Learning, DT11 Sensor, PH Sensor, Cloud, Wi-fi Module, Artificial Neural
Network, Arduino, Deep Neural Networks
166
ISBN: 979-8-3503-4821-7
PAPER: EC592
Critical Review for Estimation of Snow Depth Using Satellite Products and Auxiliary Parameters
Sanjima Rana, Priestly Shan, Kuldeep Kumar
[email protected]
Abstract— Estimating snow depth is a crucial component and has a big impact on climate change, the
water cycle, the energy balance between the surface of the Earth and the atmosphere, and the ecological
environment. The high-altitude Hindu Kush (HK) regions are completely covered with snow during the
winter period and are responsible for climate change and hazard studies. Therefore, snow depth (SD)
estimation is an important field of study in the HK region. As a result, remote sensing is the primary source
of information on large-scale SD. Microwave SD remote sensing research has advanced for more than 30
years, starting with the launch of the Nimbus-7 satellite. Researchers have made strides in a variety of
areas, including electromagnetic scattering and emission modeling, ground and airborne experiments, and
inversion algorithms for a future global high-resolution SD remote sensing mission. This paper will provide
an overview of the research and developments in the fields of SD inversion algorithms using passive
microwave data and electromagnetic scattering/emission modeling over snow-covered terrain.
Keywords— snow depth, passive microwave data, and snow cover area
PAPER: EC612
101 sequence detector using QCA technology
Ritwik Mishra, Braj Kishan Singh, Himanshu Singh, Vijay Kumar Sharma
[email protected]
Abstract - This paper mainly focuses on the implementation of finite state machine (FSM) using a novel
Quantum-dot Cellular Automata (QCA) technology. More specifically, authors have designed a 101
sequence detector, which recognizes the pattern 101 as input and sets the output flag after receiving 101
sequences. This is the first attempt when a FSM is designed using this new robust technology and authors
didn't get any previous work on this topic. The whole analysis and designing of this topic is done with the
help of QCA Designer. First, the analysis of 101 FSM is done that concludes its state tables and state
diagrams and ultimately come up with state equations. The next task is to implement the FSM’s equations
in QCA technology. To implement the equations, majority gates, inverters, and D flip-flops are needed.
Authors have used QCA designer-E software for the implementation of circuits and the calculation of
energy simulation results.
Keywords: QCA, Sequence detector, Nanotechnology, Energy dissipation, FSM
PAPER: EC617
Graphene based Rectangular Wideband Antenna for High Speed and Short Distance Wireless
Applications
S Jagan Mohan Rao, N Suguna, R Saravanakumar, V L N Phani Ponnapalli, Alamelu alias Rajasree
S, Ch Murali Krishna
[email protected]
Abstract—Proposed wide band terahertz (THz) antenna with partial ground plane incorporating on
polyimide substrate material consisting graphene resonant elements for high speed and short distance
wireless applications, which operates on the frequency spectrum of 0.45 to 1.73 THz. The antenna is
167
ISBN: 979-8-3503-4821-7
investigated for the multiple electrical parameters such as impedance bandwidth (IBW), voltage standing
wave ratio (VSWR), impedance, group delay and gain characteristics. The corresponding IBW, VSWR,
fractional bandwidth (FBW) and gain values of the proposed antenna are 1.28 THz, 1.28, 117.43% and
3.62dB, designed within the compact size of 150µm x 150 µm x 9.6 µm.
Keywords - Wideband, Terahertz, Compact size, Fractional bandwidth, High speed & short distance
communications.
PAPER: EC669
A Survey on Spectrum Sensing Using Energy Detection for 4G Waveform
Amarjeet Pal Cheema, Arun Kumar, Ahalya Gonuguntla, P. Anudeep Kumar Reddy,
Siddareddygari Gnapika, Uppaluru Naga Nanditha
[email protected]
Abstract— Spectrum Sensing (SS) is the main and most important thing Cognitive Radio (CR) does to
find unused bandwidth. This paper talks about the pros and potential of cognitive radio and gives a broad
description of how it works. Spectrum sensing based on energy detection (ED)is popular and has several
advantages. A characteristic of a transmission signal, a transmission channel characteristic, or even the
modulation technique. These articles give a simulation overview of ED over the ricin channel. The most
important things that are looked at and simulated are the probability of detection (Pd), the probability of a
false alarm (Pfa), and the false alarm rate (BER) of time- dominant energy detection. As the number of
samples increases, the framework gets more complicated and the throughput goes down.
Keywords— Cognitive Radio (CR), Energy Detection (ED), spectrum sensing, wireless communication
PAPER: EC670
Review on Animal Detection for Farmlands using Image Processing and IoT
Amarjeet Pal Cheema, Bhawana Khokher, Nikhil Jha, Nirmit Verma, Nikitha, Rakshita
[email protected]
Abstract—Monitoring agricultural land for animal infiltration is the objective of a crop protection system.
India is a nation where the majority of its economy relies on agriculture, and where farmers form the
country's backbone. The country's agricultural industry has been afflicted by a number of issues, with the
entry of unwelcome animals being one of the most significant risks farmers faces. This results in crop
damage, which leads to a decline in productivity. Traditional tactics consisting of maintaining a watch and
observing the area put a man in risk during an incursion. Therefore, a device that automatically monitors
the area without harming humans or intruders is required to avoid conflicts between humans and animals.
This suggested system combines automated detection and repellant technology with the internet of things,
which informs the appropriate authority, such as the farmer or forest authorities, when an intruder is
identified and employs the harmless impact of sound to repel the invader. It employs wireless sensors, an
image processing unit, and a camera that captures continually using GSM to warn the individual. This
survey describes the system's components and implementation techniques.
Keywords— farmlands, gsm module, alerts, defence, farmer
168
ISBN: 979-8-3503-4821-7
PAPER: EC671
A review: Secure and Efficient Data Aggregation in Wireless Sensor Networks using the Internet of
Things
Mr. Mohammed Sirajudheen A, Dr. D. Jasmine David
[email protected]
Abstract-- Data aggregation is an energy-efficient strategy in wireless sensor networks (WSNs). Due to
high node density in sensor networks. This may be avoided by employing the data aggregation method
when routing packets from source nodes to the base station. To ensure data integrity and availability in the
cloud and IoT storage system, users need to verify remote data integrity. In this publication, many ways
have been discussed to eliminate data, especially in data aggregation through security and efficiency. Data
aggregation is one of the key ways to eliminate data redundancy and energy efficiency enhancement; it
boosts the Wireless Sensor Networks' lifetime. Securing data aggregation aims to secure data aggregation,
one of the essential primitives of sensor networks for communication efficiency, from security threats such
as eavesdropping and forging aggregated findings. Considering the key elements and obstacles of data
aggregation in WSN, an evaluation of several types of data aggregation techniques and protocols were
offered in this approach with secure and efficient verification. In WSNs, data aggregation is a process of
gathering and integrating usable information in a certain area of interest. The efficacy of the communication
across nodes relies on the data aggregation strategy being employed.
Keywords-- Data aggregation, Wireless sensor network, Internet of things, Securing, Protocols.
PAPER: EC676
Design of Metamaterial Inspired Antenna Using Machine Learning and Optimization Techniques
R. Gowri Manohari, K. Kavitha, R. Arun
[email protected]
Abstract— Now-a-days, the antenna design process is facing challenges and it involves compromise
between antenna structure including dimensions and overall performance parameters such as Gain,
Bandwidth, Return Loss, and Voltage Standing Wave Ratio (VSWR). Most of the antennas are designed
by using optimization methods which focus only on antenna size reduction, but the overall performance
optimization of antenna is not much focused. Because satisfying many different contradictory performance
requirements are difficult in antenna optimization. Hence, this work analyzes the formulation of a better
objective function for satisfying different performance requirements of a metamaterial antenna. In this
work, an antenna model is developed using Artificial Neural Network (ANN) and optimization of the same
is done by using suitable objective function for the antenna model. To analyze the right choice of objective
function four different single objective functions with constraints are formulated and its effectiveness has
been explored. Further, the popular Evolutionary Algorithm Particle Swarm Optimization (PSO) method
is employed to optimize the ANN based metamaterial antenna parameters. The minimizing objective
function with constraints shows better results compared to maximizing objective function. Also, it is
inferred that more than one performance requirements can be specified as constraints. This kind of
optimization helps to satisfy the overall performance of the antenna.
Keywords— Metamaterial Antenna; Antenna design optimization; Machine learning; PSO; ANN.
169
ISBN: 979-8-3503-4821-7
PAPER: EC677
Intelligence Harmless Animal Avoider in Agriculture Farm
P.Stephen Charles, S.Dharmalingam, M.Baskaran, P.S.Manoharan, G.R.Hemanth
[email protected]
Abstract— It is very important issue in India to solve the problem of wild animals attacking agriculture
fields. This paper develops a prototype animals avoider model to surveillance the field 24x7 which is not
possible for a human being to protect the crops from damage caused by animal in addition to divert the
animal without harming them. The proposed system consists of Arduino, PIR motion sensor, LED flash
lights, speaker and acoustic sound gun. Whenever the animals are detected by the PIR motion sensor,
Arduino will be activated and speaker will sound to threatening the animals, LED lights will flash and send
SMS and photo to the farmers within few seconds of the detection of the animals. This system ensures the
safety of farm and decreases the loss of farmers as well as protects the animals.
Keywords— Animal avoider, Microcontroller, PIR Sensor, Sound gun, Agriculture farm.
PAPER: EC679
FILTENNA WITH DEFECTED GROUND STRUCTURE FOR 5G APPLICATION
Sai Haneesha Tanguturu, G. Soundarya, R. Kiruthika, Dr.N. Gunavathi
[email protected]
Abstract— To suppress radiations from undesired frequencies, a novel proximity feed filtenna is to be
designed. The proposed model operates at 5.6 GHz in the 5G midband frequency range. A defected ground
structure (DGS) is introduced to reject the radiation from unwanted frequency, a rectangular ring slot has
been incorporated on the partial ground plane parallel to the rectangular patch, resulting in a reduction in
patch element area compared to conventional designs. The proposed concept has strong implementation
potential for 5G applications.
Keywords—Proximity feed, Defected ground structure, Filtenna, 5G Mid band frequency, Partial ground.
PAPER: EC682
Binary bat swarm algorithm for thinning antenna elements in linear array
Sahiti Vankayalapati, Kailash Kavala , Bhargav Prasanna Kumar Anna, Mahender reddy
G,Lakshman P, Shekar R
[email protected]
Abstract— Binary bat swarm optimization algorithm (BBSO) has been utilized to the synthesis of a large
antenna array having linear geometry. Thinning technique is adopted to synthesize the large array. This
thinning is difficult to achieve using traditional stochastic algorithms as they fall short in optimizing large
number of discrete parameters. The work in the paper depicts how efficiently a large, optimized array can
be derived by using thinning. The BBSO is utilized to derive the optimal positions of the periodic array to
achieve a peak sidelobe level (PSLL) below -20dB. Reduced PSLL is requisite for modern communications
systems that help them in avoiding interference from unwanted directions. A linear array having 40
elements is considered for synthesis.
Keywords—binary, bat, optimization, thinning, array
170
ISBN: 979-8-3503-4821-7
PAPER: EC683
Interfacing Kera’s deep learning technique for Real-Time Age and Gender Prediction
Sahiti Vankayalapati, RShekar, Y. Rupa Sindhu, M. Manvith Reddy, P. Sasidhar Reddy, Ali Baig
Mohammad
[email protected]
Abstract-- The 21st century has seen the humans achieved many unimaginable achievements in the world
of technology. This technology can be used in a different way as individual desire. It can be used to identify
a person’s age and also identify the gender of a particular person by detecting face of the person with help
of a camera image or even with a real time video. In this research paper, a complete procedure, what are
the different methodologies and algorithms that can be used, and which one is more accurate method to be
used will be shown. The aim of this work is designing as well developing model which can detect faces
thereby predict the age and gender from real time images. This model can be used in police surveillance,
policing to dating applications and much more. Deep learning techniques are used in designing this model.
Convolution neural network and computer vision is also being used in this paper to adapt to real time
processing. Convolutional neural network along with the computer vision and keras is used in training the
model. After extensive testing and training, accuracy of 85 percent is achieved for the developed model.
This model can be broadly used in airports, age restricted areas or buildings and can be used in restricting
the alcohol access to the under aged people.
Keywords--- Convolutional neural network; deep learning; face recognition; Feature extraction; keras
algorithm.
PAPER: EC693
Parkinson’s disease classification using Machine Learning techniques
Meenakshi K, D.Kishore
[email protected]
Abstract—Parkinson disease is one of the dangerous disease that causes mortality at advanced stage. It is
estimated that 9.6 million people are affected through out the world. These statistics dictate that this disease
requires urgent attention which causes failure of motor system in humans causing the mental and physical
disabilities. Therefore, In this work, machine learning algorithms are employed to aid the practitioner in
medical diagnosis.
Keywords—Parkinson Disease, Feature extraction, Machine learning algorithms, Accuracy, Receiver
operating characterestics.
PAPER: EC698
Automatic Shopping Cart Using RFID and NodeMCU
Jalalu Guntur, S. Srinivasulu Raju, M. Srikanth, N Surya Seshu Kumar
[email protected]
Abstract— A culturally perceived imaginative item will coordinate the solace, comfort, and adequacy in
regular daily existence. The Internet of Things, or IoT, is an organization of interconnected enlisting
gadgets, mechanical and automated machines, things, animals, or people with exceptional identifiers
(UIDs) and the capacity to move information across an association without the requirement for human-to-
human or human-to-PC cooperation. The Internet of Things, or IoT, is an organization of interconnected
171
ISBN: 979-8-3503-4821-7
enrolling gadgets, mechanical and mechanized machines, things, animals, or people with novel identifiers
(UIDs) and the capacity to move information across an association without the requirement for human-to-
human or human-to-PC participation .In metro urban areas, purchasing and shopping in monstrous
shopping centers has turned into an everyday action. The Internet of Items (IoT) alludes to the most
common way of connecting everything on the planet to the web. Individuals buy a wide scope of things
and spot them in their shopping baskets. Following the culmination of the buy, one should continue to the
charging counter to make an installment, which is both tedious and disappointing. The significant objective
in making this model was to decrease human work, dispense with lines, and abbreviate charging time for
clients. The model is comprised of RFID labels that are utilized to distinguish items, as well as a RFID per
user that filters items as they are set in the truck and shows the outcomes on a LCD show. The NodeMCU
will save the data from the RFID labels. NodeMCU is utilized rather than a WiFi module. The information
is then moved to the server from the charging counter. The client just requirements to take care of the bill
at the charging counter. The objective of this undertaking is to assemble and make sense of how the Internet
of Things (IoT) can be utilized in a shrewd shopping basket.
Keywords: RFID, ESP8266NodeMCU, Internet of Things (IoT), Smart Shopping cart, WiFi
PAPER: EC700
A Fork Shaped UWB Antenna with WLAN/WiMAX/ X Band Rejection Characteristics
D. Sreenivasa Rao, Shailendra Kumar Mishra, Jetti Chandrasekhar Rao, Mogadala Vinod Kumar,
M. Vasuja Devi Department of EC
[email protected]
Abstract— A novel fork shaped patch antenna for UWB (Ultra-wide Band) products is devised on 27 X 27
X 1.6 mm3 FR4 epoxy substrate. Slots are inserted in the partial ground plane to improve the bandwidth
of the antenna. Three band stop functions are accomplished by carving slits in the ground plane, feed line,
and radiating element at 5.15–5.85 GHz (WLAN), 3900–4200 MHz (C-band satellite communication),
3.3–3.75 GHz (Wi-MAX), and 7900–8400 MHz (X-band satellite communication frequency). All slots are
optimized by parametric analysis. The proposed antenna showing good impedance bandwidth stretching
from 3100 to 10600 MHz. The prototype shows good radiation characteristics and observed good match
between simulated and experimental results.
Keywords— Ultra-Wideband, Slots, Stub, Band notched
PAPER: EC701
Classification And Detection Of Skin Cancer Using Deep Learning Methods
Chivukula Venkata Sanjana, S. Srinivasulu Raju, Munji Anusha, Boyapati Sivani
[email protected]
Abstract—The types of Skin Diseases are almost universal around the world because they are caused by
genetic and environmental factors. Skin cancer has a very high mortality rate but is usually curable with
minor surgical intervention [1]. Diagnosis of melanoma begins with clinical screening followed by
dermatoscopy and histopathological examination. Melanoma skin cancer is highly curable if treated early
buy detection is hard, the medical procedures are too long, and the doctor’s diagnosis cannot be 100%
accurate. High speed cameras have 60- 80% high accuracy stated by International Skin given Imaging
Collaboration (ISIC). The aim of this project is to determine an accurate prediction of skin cancer and
172
ISBN: 979-8-3503-4821-7
classify skin cancer as malignant or non-malignant. A quick and accurate diagnosis can be of great benefit
to doctors and patients. In many areas of medicine, the performance of recent models based on deep
learning approaches or even exceeds that of human experts. In this study, to help of dermatologists improve
the efficiency of skin cancer analysis, we will use Convolutional Neural Networks (CNN) with Convnet
model to complete the recognition of skin lesions images. This project aims to shorten the limitations by
providing the predictive model using CNN to classify 9 types of skin cancer lesions from outlier lesion
images. Proposal models are trained and evaluated using international standard dataset.
Keywords-Deep learning, Convolutional Neural Network (CNN); classification, skincancer. Melanoma;
Non-Melanoma;
PAPER: EC706
Design and Implementation of Fuzzy Logic Controller for Boiler Temperature Control using
LabView
K Niharika, S. Srinivasulu Raju, Y Sujan, K. Sai Tarun, B. Teja
[email protected]
Abstract—This paperwork aims design, implementation of fuzzy logic controller for controlling of water
in the tank using LABVIEW software. This system mainly contains the single tank, Resistance
Temperature Detector sensor and myrio. The signal from the temperature sensor is transmitted to
LABVIEW software via myrio interface, connected to the system. Fuzzy logic controller (FLC) algorithm
is designed using knowledge-based inference engine, which can be used to replace the conventional
controllers tuning parameters using analytical equations. The performing of fuzzy logic controller for
controlling the temperature has been explored in detail. It provides a improved response, and quickly tracks
the set point. FLC functions well with the system involved uncertainties, noise at sensory signal. This
developed FLC could also be used to measure several variables like temperature at industrial applications.
Keywords --FLC, LABVIEW, PID, Performance Analysis.
PAPER: EC718
Design and analyzing of hexagon-shaped microstrip patch antenna for biomedical applications
Elangovan. S, Gokul raj. R, Vethaprasath. M, Mahendran K
[email protected]
Abstract-- The recent development of high speed wireless communication systems has made tremendous
research efforts towards increasing the capacity of wireless systems in terms of large bandwidth, data rate
and high speed communication at lower cost. Microstrip patch antennas are a special class of antennas that
are partially or entirely made out of microstrip patch materials, in contrast to conventional antennas, which
consist of rigid materials. A compact low-profile multi band microstrip patch is implemented for
biomedical applications composing a filtering antenna for higher bandwidth composed of electrically
conductive substrate, electro microstrip patch and applied for the radiating and grounding parts, and
dielectric materials for the insulating parts of the antenna.The proposed microstrip patch antenna
composing of a slotted microstrip patch antenna are divided into electrically conductive fabrics, denoted
electro Microstrip patch and defected mono pole ground structure, and dielectric materials for the insulating
parts of the antenna. The proposed high bandwidth multi band microstrip patch antenna achieve ISM band
2.4Ghz frequency.
Keywords: hexagonal shaped antenna, microstrip patch antenna, ISM band, antenna design, antenna design,
2.4Ghz frequency antenna, monopole ground structure, 2.4Ghz frequency, gain of 3dB, bio medical
applications.
173
ISBN: 979-8-3503-4821-7
PAPER: EC719
Reconfigurable Hardware Implementation of CNN Architecture using Zerobypass Multiplier
SAKTHI.K, ABISHEK.D, ARUN KUMAR.M
[email protected]
ABSTRACT-- The current state-of-the-art in image recognition, segmentation, and localization algorithms
has reached an extremely high level of accuracy thanks to the development of deep neural networks and
their use in computer vision problems. Specifically, Convolutional Neu- ral Networks (CNNs) have
reached human-level performance in image classification and detection. CNN must be executed on a
portable, low-cost, and low-power-consuming device for the object classification/detection use cases.
However, real-time execution of CNN based applications is hindered by these devices' limited computing
resources and low on-board memory storage capability. In this dissertation, we introduce hardware-
efficient algorithms for performing complex computations in CNN at low cost and with minimal power
consumption. Additionally, we provide efficient VLSI architectures based on systolic arrays for creating
high-performance devices for running CNN. The limitations of traditional CNN for detecting occluded
objects are also outlined in this thesis. We propose an improved CNN with self-feedback layers and present
an algorithm to increase the detection accuracy of the hidden objects. Improved accuracy in detecting
hidden objects is found when the enhanced CNN is compared to the gold standard dataset. In addition,
enhanced CNN requires a much larger number of computations to execute than regular CNN. We introduce
a low-cost, low-power VLSI architecture design for efficient hardware execution of improved CNN.
PAPER: EC720
A Traffic Road Sign Recognition System using Raspberry pi with Open CV Libraries Based On
Embedded Technologies
AJAY SRIRAM.D, RADHAKRISHNAN.P, RAJA.E, PRAVEEN.M
[email protected]
Abstract— The use of computer systems for the purpose of image processing, which is also known as
digital image processing, has applications in a variety of sectors, some of which include geology,
oceanography, medicine, and space exploration to name a few. The major objective of this piece is to
provide examples of different ways in which image processing techniques may be used even in
environments with limited computational resources. In particular, we developed a traffic sign recognition
system that is capable of reading and interpreting speed signs. This system is based on embedded
technologies, and it was constructed by us. This research covers a wide range of topics, some of which are
the characteristics of speed limit signs, the requirements and challenges of constructing an embedded real-
time base system, and the methods for manipulating numbers that make use of image processing techniques
that are based on form and dimension analysis. The study also provides a presentation of, and a
demonstration for, several classification and recognition strategies. This research brings to light a number
of difficulties with the consistency of color identification as a consequence of lighting circumstances; as a
result, avoiding the usage of color models may result in a more effective solution than using color models.
In addition to this, color analysis plays a particularly significant role in a broad range of other applications
for the identification of road signs. These applications may be found all over the world. For the most part,
lightweight solutions were selected for this project since the real-time applications that could be used as
well as the capabilities of the Raspberry Pi were severely limited. The Raspberry Pi is the primary focus of
concentration for the implementation as a whole. The Raspberry Pi is able to connect sensors, databases,
and the results of image processing, in addition to carrying out actions on devices that are external to itself.
174
ISBN: 979-8-3503-4821-7
PAPER: EC721
REVIEW ON CARDIOVASCULAR DISEASE PREDICTION USING MACHINE LEARNING
ALGORITHM
Adhishayaa P, V Gomathi, V Mahendran K
[email protected]
Abstract— cardiovascular diseases (CVD) are the one amongst foremost reasons for death around the
globe over the past decades. Early detection of heart diseases and continuous monitoring reduces the death
rate. The heart is responsible for blood circulation in the entire human body. Heart attack is happening
because of coronary artery disease (CAD) and chronic heart failure (CHF). Angiography is one of the most
common medical diagnosis procedures for heart disease. Nowadays, researchers are concentrating more on
this cardiovascular disease area. Based on Machine Learning (ML) the diagnostic procedure is developed.
ML provides more accurate and efficient diagnostic results that helps to diagnose the CVD in the initial
stage. This paper comprises numerous Machine Algorithms for the prediction of cardiovascular diseases
and compares the accuracy of those ML algorithms.
Keywords: cardiovascular diseases, Random forests, K- closest neighbors, Support Vector Machines
PAPER: EC723
Efficiently Enhancing Energy by means of LEACH protocol for a Wireless Sensor Networks
Raji Pandurangan, Jagadeeban T, Girish M, Tholkapian K
[email protected]
Abstract — In potentially dangerous situations, Wireless Sensor Networks (WSNs) need to be touted as
means of a viable method for consistently monitoring the civic and military environments. Typically,
meaningful field data is gathered using sensor webs, which are composed of nodes with limited power and
wireless connections. Improved energy efficiency is a major concern for WSN protocols due to these
factors. The Low Energy Adaptive Clustering Hierarchy (LEACH) protocol proposes a method in which
sensor nodes are clustered and data is fused prior to being sent to the base station (BS). The SD-LEACH
protocol is suggested because certain applications are hampered by WSN drawbacks such as poor battery
life, short lifespan, restricted sensor deployment range, and energy consumption. In this paper, we present
Secondary Data Transmission LEACH (SD-LEACH), a novel method for enhancing the LEACH protocol
for usage on sensor networks. SD-LEACH decreases packet delivery time to increase network lifespan,
speeding up data transmission.
Keywords: WSN, LEACH, SD-LEACH, , Energy, PDR, Network Lifetime.
PAPER: EC724
Smart solid waste collection for a sustainable smart city usingArduino Uno
S Ramya, T K Srinivasan, P Sai Pavan, M V Hari Haran, G Vishnuvardhan, D Lingaraja, G Dinesh
Ram
[email protected]
Abstract— The globe over, garbage disposal has recently been a serious cause of concern. There is an
excessive amount of waste produced and disposed of in ways that are bad for the environment. There are
175
ISBN: 979-8-3503-4821-7
many different sources of garbage, including residential and industrial waste. People frequently focus on
industrial waste because they don't think to separate waste into its basic streams, which is the source of
commercial waste. The smart dustbin is a simple and cost-effective solution for a segregation system for
small and medium-sized firms or industries that supply products directly for processing. The appearance
distinguishes between metallic and non-metallic trash, such as Dry, Wet, and Metals. The Arduino UNO
Micro is used to control the circuit.
Keywords— smart dustbin; segregation waste; garbage disposal
PAPER: EC725
A Novel Framework on QoS in IoT Applications for Improvising Adaptability and Distributiveness
Mansoor CMM, Vishnupriya G, Anand A, Vijayakumar S, Kumaran G, Samuthira Pandi V
[email protected]
Abstract— The Internet of Things, or IoT, is a broad term referring to a pattern of technology that can
detect, compute, transmit, and store the data that must be exchanged and stored in end user applications
and systems. IoT also performs the functions of gathering the information, processing the information,
analyzing the data. Due to wider range of scope and increase in number of IoT devices, it has become a
critical challenge in meeting the demands of Quality of Service (QoS) in IoT devices. So there is a great
need to access various parameters of QoS based on many perspectives, likely user specific, operator and
network specific, depending on communication resources, edges nodes and all other enabling technologies.
Keywords—Quality of service; Deep neural network; Artificial neural network; Application layer; Network
layer
PAPER: EC727
A Novel Interference Mitigation Technique and Signal Strength Prediction in future Wireless
Communication
Helina Rajini Suresh, Beulah Jackson, Dhanalakshmi K M, Anand A, Ramkumar S, Asha.A
[email protected]
Abstract— The modern cellular systems are the key enabler for the development of recent wireless
communication systems. So a new concept of frequency reuse has been developed over spatial domain in
cellular systems. This allows a huge number of people to interact spontaneously. The fast growth of
wireless technology has raised data communication needs. As of now Long Term Evolution (LTE) is
considered as the best encouraging standard for future mobile broadband services. Thus mobility
management is the process to trace the Mobile Subscribers Unit (MSU) point for handing over the data
packets and retain a MSU link as it continuously changes its point of connection. This paper investigates
the performance of STBC backed by correlated frequency selective channels for downlink
communications. Using the STBC-based MIMO profile and the Maximum Likelihood approach at MS, the
required user information may be estimated. By lowering the size of the mobile unit employing DL
communication, the suggested system achieves superior performance.
Keywords—Low density parity check codes; Mobile station; Bit error rate; Base station; Orthogonal
frequency division multiplexing.
176
ISBN: 979-8-3503-4821-7
PAPER: EC728
Block Chain Based Identity Authentication System for a Wireless Sensor Networks
Naresh Kumar Thapa K, Vishnupriya G, Manisha G, Omar S Saleh, Lakshmibai T, Uthayakumar
GS
[email protected]
Abstract— In the present IoT context, typical focused security validation approaches are unable to offer
security verification due to the restricted performance and flexibility of the devices. These issues are
addressed by IoT-Chain, an Internet of Things (IoT) security validation system built on the Hyperledger
Fabric block chain architecture, which offers superior security confirmation. These three forms of chain
codes are all included in the architecture. It is the entry code that is responsible for running the test to
determine whether or not a client is healthy. An asset information URL may be retrieved via a device code,
whereas a method code indicates the access control mechanism for a particular client. IoT-Chain offers
dynamic security validation of the Internet of Things by integrating access control with block chain
technologies. IoT-Chain can sustain high throughput and really achieve consensus in a distributed system,
according to preliminary studies.
Keywords—Quality of service; Recurrent nueral network; Inductive matrix completion; Attention
mechanism; encoder decoder architecture.
PAPER: EC729
Digital Filter Design on High speed Communication with Low Power Criteria
Rajalakshmi R, Vishnupriya G, Sudharsanan R, Navaneethan S, Vijayakumari P, Karthikeyan M
V
Abstract— The growing popularity of multi standard wireless communications has created a significant
need for DSP. Signal filtering is a common DSP procedure. Digital filter applications are found in data
communication, speech processing, audio processing, biomedical signal processing, image processing,
instrumentation and control. Implementation of digital filters on Field Programmable Gate Array (FPGA)
has gained importance in recent years due to flexible reconfigurable architectures. In this article, the design
and implementation of digital filters for communication applications, such as SDR and noise cancellers,
are discussed. These filters may also reduce noise. Synthesis parameters like resource utilization and
maximum frequency of operation are compared with the existing methods. The implementation results
show that the proposed methods offer less resource utilization, particularly in terms of the number of
multipliers which are more power consuming parts in the FPGA, which inherently indicate low power
implementation of the developed methods.
Keywords—Numerically controlled oscillator; Programmable digital signal processor; Digital down
converter; Adaptive noise cancellation; Gradient vector computation.
PAPER: EC735
System for Real-Time Monitoring and Reporting of the Weather Based on the Internet of Things
Sakthi K, Ashwin Kumar R, Chidambaram V, Gokul Saran K
[email protected]
Abstract — The findings of this research point to a cutting-edge technology that has the potential to be
utilized to keep track of local climate conditions and to make the information that it generates available in
any area of the world. The technology that makes this possible is known as the Internet of Things (IoT),
and it is a cutting-edge method that is both efficient and effective for linking everything in the world of
things together in a network. It is a cutting-edge approach that is known as a means that is cutting-edge.
177
ISBN: 979-8-3503-4821-7
There is a chance that this area holds items like electronic devices, sensors, and maybe even the electronic
components from vehicles. By using several sensors, the system is able to keep track of and make
adjustments to a number of characteristics of the environment, including the temperature, the relative
humidity, and the levels of carbon monoxide. After the information has been provided, the data from the
sensors are then shown as graphical statistics on a web page that is able to receive it after it has been sent.
Anyone in the globe who has access to the internet may see the most current data that was produced by the
system that has been in place for some time.
Keywords : IOT ,technology ,electronic devices ,sensors ,temperature ,carbon monoxide ,cutting-edge
PAPER: EC740
Coal Miners Activity Sensing using IoT
Senthilkumar M, Govindaraj T, Pradeepkumar G, Manimuthu Ayyannan, Gopu A P, Neelam
Sanjeev Kumar
[email protected]
Abstract— The importance of coal mining has grown in tandem with the rise in electricity demand.
Inadequate protection for miners breeds public hostility toward the mining industry. Mine explosions are
typically caused by the accumulation of gases such as methane. Explosions, injuries, and even death can
occur as a result of exposure to hazardous gases in underground mines. In light of these considerations, we
created a framework that uses Arduino technology to monitor dangerous substances, abnormal temperature
conditions, the human heart rate, and blood oxygen concentration. The enhanced safety features of our
system significantly increased coal miners' awareness of the risks by warning them of them. To monitor
the conditions of a coal mine, our system employs a circuit that collects data on temperature, heart rate, gas
pressure, and oxygen level. In the event of an emergency, the buzzers will alert both miners and those
working outside the mine, giving everyone enough time to flee.
Keywords— Coal Mining, IoT, Activity Sensing, Arduino Node MCU
PAPER: EC741
Development of IoT based Solar Water Heater Controller
Lekashri S, Manimuthu Ayyannan, Murugesan A, Meenatchi K, Rajesh Kumar B, Neelam Sanjeev
Kumar
Abstract— The paper analyzing and improving the efficiency of a solar water heating (SWH) system. The
obtained IoT data is used to gain insight into the system's operating and environmental conditions, to
identify areas where individual SWH component performance can be improved, and to design control
strategies to ensure the system runs smoothly. We conduct a thorough energy audit to analyze the
performance of subsystems as well as the overall system using Internet of Things (IoT) data such as water
flow, heat pump and water pump working schedules, solar availability, and electricity consumption. The
results of this data analysis are used to guide the development of control strategies for the SWH's numerous
subsystems, with the goal of improving the system overall and lowering operating costs. An experiment
carried out at the fictitious SWH location validates the efficacy of the proposed mitigation techniques.
Keywords— Solar Water Heating System, IoT, Controller
178
ISBN: 979-8-3503-4821-7
PAPER: EC742
VIRTUAL MOUSE
Raji Pandurangan, P V Praharsha Reddy, P Sathvik Reddy, V Sai Gurukrishna
[email protected]
Abstract-- A great example of HCI innovation is the mouse. A wireless or Bluetooth mouse still requires
electricity in order to function, and it is not totally power-free since it requires some kind of energy to
function and a dongle to connect it to a computer. The AI virtual mouse solution we presented gets over
this problem by utilising a camera, either a built-in or a webcam, to record our hand movements and a
convolutional neural network to identify the tips of our fingers. The algorithm employed in the proposed
system is a machine learning algorithm. The computer's left-click and pointer functions may be
manipulated without a hardware mouse by using hand gestures to operate the machine. In order to track the
hands and react to their movements, the system relies on deep learning. Therefore, our suggested method
will prevent the spread of covid-19 by doing away with the need for a physical mouse and reliance on
gadgets to regulate the position of the pointer
PAPER: EC758
Image Reflection Removal using Iterative Boost Convolutional LSTM Network through Feature
Loss
Harshini Daggu, Chandana Chiluka, Mothe Deekshith Reddy, Dr. D Praveen Kumar
[email protected]
Abstract- Unwanted reflections commonly degrade the clarity of images taken via glass windows. The
elimination of that particular reflection is essential for enhancing image quality. Due to the poorly worded
challenge, many solutions require many input photographs, which in practice looks to be difficult. Users
using the alternate solution must take numerous images from various perspectives. To decrease this
problem, we thought about employing the cascaded network that is a series network model for the
transmission and reflection layers. A straightforward approach for a cascade model on IRR is to educate
one network to produce a forecasted transmission that act because the auxiliary data of the following
network by continuing the method iteratively utilizing subnetworks that enhance the forecast. After passing
through all the layers of network, there will be some losses present in an image which produces blurry
images. In order to remove losses, the image is moved to the initial step.
Keywords—Reflection Removal, LSTM, IBCLN, Gradients, Feature Loss.
PAPER: EC776
Design of T and Y Micro-Channel through Multi physics COSMOL
Saranya B, Abinaya V, Nirosha R
Abstract— Miniaturization of products is the recent trend in the Global market where almost every product
needs miniaturized size with stabilized performance. the development of micro systems are derived from
electro mechanical system implementations for the development of Fluid Mechanics. The represented
system is focused on developing microfluidic devices applicable for medical divine biological systems
towards medical products etc. the proposed system is focused on developing Yshaped microchannel and Y
179
ISBN: 979-8-3503-4821-7
shaped microchannel provided with testing factors on pressure velocity concentration towards the channel
developed here using are small multiphysics to. The presented approach considers the microchannel
development and cleanly focuses on the performance of both the channels using RENOLDS number. in
which many biological analyzers and drug delivery systems microchannels aur implemented. The
performance of microchannel is important in which the micro Fluids mixture various microfluidic mixing
techniques are implemented where the performance of vi channel based microfluidic mixing and T-Shaped
based microfluidic mixing is stimulated here. The present system has a high Renolds number for Y- channel
mixing compared with T-channel mixing.
Keywords— Micro fluids, Micro-mixers, Micro-electronics, Microchannels, thermocouple.
PAPER: EC777
EMOTION ANALYSIS OF TWEETS
D.Durga Prasad, Pardhu Guttula, Tangellapalli Sai Ramya Manasa, VemulapalliBavya Sri, Thote
Prashanth, Pitani Niranjan Sai
Abstract—As technology isdeveloping day by day messages, social media platforms became a platform
of representing person’s life, thoughts and theirbehaviour. We all know this developing technology has
more freedom to share their thoughts, personal feelings, what they are feeling at the current moment within
a single second, within a single sentence on social media. There will be plenty of data available to know
about a personbehavior and this data can be used for detecting various emotions. Our project Emotion
Analysis of tweets itself says that to detect a particular emotion held by a tweet. That is done by analyzing
expressed opinions, images, sentiments, and other activities. In our case we took mainly posted opinions
or tweets as input to implement our NLP (Natural Language Processing) model. This project is about
constructing machine learning models to find which model is predicting emotion of tweets with highest
accuracy.
Keywords—Twitter, Sentiment, Social media, natural language processing
PAPER: EC778
Optical Imaging for Telephoto Applications Featuring a Lens Design Optimized For Thermal
Compensation
Srinivasan T.S, Vivek Maik, Samiappan Dhanalakshmi
[email protected]
Abstract—Temperature fluctuations in optical imaging systems often affect picture resolution. The quality
of the optical element varies due to the high temperature, resulting in focus shift. An optical system for
severe temperatures ranging from -40° to +80° is examined in this paper. To investigate physical and
geometrical aberration, a five-element telephoto lens system with the following specifications is used: The
entrance pupil diameter (EPD) is 25 mm, the visible wavelength is 486, 587, 656 nm, the field of angle is
0, 7, 10, the F- number is 4, and the semi field of view (FOV) is 10. This preliminary system performance
measured and observed picture quality using modulation transfer function (MTF) and spot diagram (SPO).
For a spatial frequency of 20 lp/mm, the MTF is 0.5, and the spot diagram reveals that the rays are not
converging in the optical axis. This system attained an MTF value of >0.8 (80% accuracy) after
optimization. Also, by using Automatic design, the error function or merit function is lowered from 1407.08
to 13.93, and further reduced to 6.77 by incorporating Aspheric surface for element surfaces 1 and 3. We
were able to develop a high resolution optical system without modifying the glass material and the overall
length of the system.
Keywords- Athermalization; lens optimization; optical design; thermal compensation; Aspheric expert
180
ISBN: 979-8-3503-4821-7
PAPER: EC782
RECONFIGURABLE MIMO ANTENNA FOR 5G APPLICATIONS
Ganesh Naidu Marrapu, R. Kiruthika, G.Soundarya, N.Gunavathi
[email protected]
Abstract— In this work, the proposed antenna operates at 2.4GHz and 3.85GHz in the 5G mid-band
frequency range. Initially, a single-element monopole antenna is developed on Rogers RT 5880. This
single-element monopole antenna and its replica with a separation of 10.6mm form 2 element MIMO. PIN
diodes are introduced to achieve frequency reconfiguration. The proposed antenna yields good isolation,
high diversity gain (DG>9.8), and a low Envelope Correlation Coefficient (ECC<0.01)
Keywords—MIMO, Reconfigurable, PIN diodes, ECC, DG
PAPER: EC783
RFID-Enabled Wireless Printing Solution Using Raspberry Pi
Hanamant B. Sale, Asha Ghodake
[email protected]
Abstract— Printers are challenging to set up correctly, and wireless printers are prohibitively expensive
for the average person. Printing documents at a business establishment necessitates mailing them or sending
them over instant messaging services, raising privacy issues. While some printers are small and stylish,
they can occupy a sizable portion of your workspace. When your desk area is congested or small, that is a
problem. One way to organize and maximize your workspace is to make your printer wireless. This allows
it to be placed anywhere you have space in your home or office. After minimizing space, the next important
consideration is security. When sending the document to someone else’s printer, it is first stored in the
system to which the printer is connected. Consequently, the document is not secure in this instance. But in
this proposed device, each user has their RFID tag, so the document will be safe even after being printed.
The system is secure, quick, practical, and simple, combined with other existing services.
Keywords- printers; documents; wireless; security; Raspberry Pi; RFID; component
PAPER : EC784
Deep Learning based Intelligent Alert System for Visually Impaired People
Krishna Dharavath, Veeramallu Tejaswi, E Sai Prasanna Kumar, Vennu Sai Pratham
[email protected]
ABSTRACT— According to a study of World Health Organization, there are about 39 million people in
the world who are completely visually impaired. These individuals often face significant challenges in
navigating and managing their daily life. They always need to be dependent on assistance from other people
or any assistance system even for a short distance. So, the main objective of this work is to design a
feedback system which acknowledges the person about the obstacles with their names. The frame in front
of the user is given as the input to the system through a wearable device. Further, it detects the obstacles in
that frame and the user is alerted with the name of the obstacle through a headphone or earphone connected.
This can be integrated with a smart cane and any assistance system which improves the safety and mobility
of the user. This is achieved through deep learning. Accuracy is the main cause for using deep learning.
181
ISBN: 979-8-3503-4821-7
Better results can be attained by deep learning models. The system has shown 96% accuracy. The two main
stages of the system, object detection & classification and the text to speech conversion are achieved
through a deep learning technique and speech synthesizer respectively.
Keywords- Visually impaired people, Deep Learning, Object detection, Object Classification, Speech
Synthesizer, Text to speech conversion
PAPER : EC785
Quadcopter Navigation using Telemetry Data

M Varun M.E.,M Naveen Joel,M Naveen,J Vishnu Haran,
[email protected]
Abstract— In this project, we demonstrate the usage of telemetry data implied on a Quadcopter. A
Quadcopter (also known as Drone or UAV) is controlled automatically by an embedded system known as
the "Flight Control System." Sensors, GPS and a telemetry system are coupled with an airframe and
computer system in this scenario. These components must perform autonomously in various missions and
situations while being supervised by a ground base station or controller that serves as a remote transmitter.
UAVs can also be effectively employed in so-called "D-cube" missions, which are more explicitly
characterized as Dangerous, Dirty, or Dull.
Keywords— GCS - Ground Control Station, RTL - Return to Launch, UAS - Unmanned Aircraft Systems,
UAV - Unmanned Aerial Vehicle, ESC - Electronic Speed Controller, BLDC – Brushless DC motor, APM
- Ardupilot Mega, SITL- Software-In-The-Loop
PAPER : EC787
Data analytics to reflect the current politics on twitter by using machine learning
Piyush Charan, N.Bindhu Madhavi, Pavitar Prakash Singh, Amedapu Srinivasa
[email protected]
Abstract— Many unstructured text messages, chats, postings, and blogs have been produced as social
media users have skyrocketed. In addition to serving as a platform for information dissemination, social
media platforms provide a natural forum for the wide dissemination of ideas and viewpoints that acquire
traction when supported by a sizable audience. The positive or negative feelings individuals have about a
person, group, or location might be reflected in how well they are known. Vast quantities of material on
social media platforms like Twitter provide political insights that may mine to understand public sentiment
and foretell election outcomes. An effort is made in this paper to mine tweets for political sentiment and
represent this extraction process as a supervised learning problem. Twitter users' attitudes toward the main
national political parties campaigning in the 2019 Indian election are analyzed, and tweets related to the
upcoming election are extracted. The sentiment analysis categorization algorithm is thus ready to forecast
the tweets' propensity to infer election outcomes. The classification model is built using Long Short Term
Memory (LSTM) and compared to traditional ML approaches.
Keywords— tweeter, Indian election, LSTM
PAPER : EC791
A Medical Assisting System using Augured Reality For Critical Analysis
Latha R , Rudhran M , Punyakotti Reddy P , Rishanth Raj R
[email protected]
Abstract— Experts frequently monitor new developments to enhance their working environment.They
frequently are the first to implement advances that allow their industry to provide a more attentive and
182
ISBN: 979-8-3503-4821-7
patient experience. As the mindful working environment has continued to grow in the computer age, a
number of developments have emerged as expectedly challenging developments. Augured reality (AR) is
swiftly becoming accessible, affordable, and sensible; as a result, it will undoubtedly be applied to medical
services to enhance clinical information utilization. Applications are currently being explored for their
position in the specialists, whether it has to do with life structures, intraoperative medical procedure, or
post-employable recuperation. The expansion of false data to at least one of the faculties that enables the
customer to conduct tasks more effectively is what is meant by AR. We suggest a framework in which
crucial information for experts is displayed on foggy glasses recollected for an AR headset and thus
integrated with this current reality view. In this project, the ongoing patient data from medical clinics
collected by sensors connected to patients receive treatment once the sensor has assessed their
characteristics, at this point it is managed, sent off by doctors using expanded reality glasses, and an alarm
is set off in case an unusual condition develops. Given the patient's present medical condition, the specialist
can take the appropriate action.
Keywords— Critical-care; analysis; medical; sensors; augured reality
PAPER : EC793
Analytical review of deep neural techniques for breast tumor detection.

Dr. Katta Rajesh Babu, Samiraju Sai Bhanu Teja, Vanga Satvik Reddy, Kattamuri Mani Deep,
Mallabattula N D R Ramesh, K Rambabu
[email protected]
Abstract— A tumor is an abnormal mass of tissue that develops when cell growth and division are
accelerated. Cancerous tumors can develop, making them extremely harmful and needing early detection
for safe excision The most difficult and crucial task in many medical image applications is tumor
detection/segmentation because it typically requires a large amount of data and information. Deep learning
techniques including CNN (convolutional neural networks), RNN (recurrent neural networks), and DBN
are employed in this paper (Deep belief network). The BRATS 2019 dataset is used in this study, and
machine learning methods such as Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and
Random Forests are compared. Performance metrics such as Dice Score Coefficient (DSC), Positive
Predictive Value (PPV), and Sensitivity are used to determine which technique performs the best overall
based on these metrics. The findings of the segmented images yielded dice scores for the CNN technique
of 0.89, PPV of 0.87, and Sensitivity of 0.90.
Keywords— Tumor, Segmentation, Machine Learning, Deep Learning.
PAPER : EC796
Regulating and monitoring IoT controlled solar power plant by ML

Ayasha siddiqua, Afsana Anjum, Sunanda Kondapalli, Dr Chamandeep Kaur
[email protected]
Abstract— The proposed approach utilizes real-time techniques to aid in regulating and monitoring of
solar power plants through the Internet of Things(IoT). Traditional PLC technology is insufficient for
remote access and monitoring of solar power station operations. Therefore, the IoT and Machine learning
(ML) are used to the administration of solar power plants. The term "Internet of Things" refers to the
integration of several physical technologies with cloud applications. Each component of the Internet of
Things (IoT) must have a certain minimum level of processing power, data security measures, and
communication capabilities. This innovative technology may be used to monitor and boost solar power
183
ISBN: 979-8-3503-4821-7
output. Since servomotors may be used to rotate solar panels in response to changes in the sun's angle, the
efficiency with which they produce electricity can be increased. Designing a photovoltaic system,
constructing the analogue circuitry for accurate voltage and current measurements, and developing a
website to provide the monitored data in an approachable graphical format are all necessary steps towards
this goal. Since the web server is located on a WAN (Wide Area Network), it is accessible from anyplace
with an Internet connection.
Keywords— Solar Power, cloud application, IoT, Machine learning, Energy
PAPER : EC808
Wireless Medical Sensor Network based Healthcare Monitoring system in Narrow Band IoT
Arunkumar.K, Thavaselvan P, Deepak Kumar.A
[email protected]
Abstract— The innovative idea enhances the functionality of the healthcare network. It swiftly examines
the narrow band Internet of Things (i-NB- IoT) systems in wireless networks, concentrating on several
fundamental characteristics for medical network descriptions. A new networking protocol called the
Internet of Things narrowband (i-NB) were created for the IoT with a low bandwidth. The critical
importance is given to monitoring the patients' heart and breathing rates while this sensor for the parameter
exchangers manages their body temperature, which is observed using a straightforward gadget. The doctor
is in charge of monitoring patient data, but our system is also offered to present the same information
alongside the identities of adjacent patients for self-defense. With the help of standard measures like i-
DD,i-ONC and i- RTT. These connected measures are able to send and receive large amounts of data,
frequently over great distances and with very high latency. For i-NB, the novel physical-layer control
channel operation is employed to transmit the scheduling opinion in the downlink and provide feedback
data in the uplink.
Keywords— Wireless Medical Sensor Networks, Internet Of Things, Narrow Band, Health Care
PAPER : EC809
Smart Library using Arduino
Sravya Valivarthi, Mrs. P. Sushma Chowdary, Lokesh Pallampati, Mubeen Mahammad, Sasidhar
Reddy Buchi
[email protected]
Abstract— Academic Institute's heart is the library. Libraries old concepts are rapidly changing nowadays.
Librarians used to use card catalogues and manually assign library work. The concept of libraries was that
they were primarily used as book storage facilities. However, this concept is evolving and giving way to
global libraries powered by computers. The library can only be managed with the help of automation. In
our project, we hope to automate the library in the most cost-effective manner possible. As a result, it would
be advantageous for small-scale libraries that cannot afford expensive automated software. We used an
Arduino and a barcode system that was relatively inexpensive.
Keywords— Library, Arduino, Automation, , Barcodes, RFID.
PAPER : EC843
Design and Implementation of IoT-based Programmable Assistive Intelligent Adaptive Pillbox

N. Bharathiraja, S. Hariprasad, M. Sakthivel, T. Deepa
[email protected]
Abstract— A pillbox is one of the most common ways for older people to take their medication. However,
this does not ensure that patients receive excessive or inappropriate doses. In this proposed work, a system
184
ISBN: 979-8-3503-4821-7
for automatically giving patients the information they require for taking their prescriptions correctly and at
the correct times. Here, RFID technology is used to provide patients with their identification. The pillbox
has three different types of tablets inside for headaches, fevers, and diarrhea. Patients have trouble with any
of the diseases they will pick caused infections in the asking question and place the figure in the heartbeat
sensor, which measures your heartbeat and temperature. These are all updated in the thingspeak IoT cloud.
After updating the values servo motor turns on, and the pill is lifted. It has become increasingly important
in the modern day to provide care for the elderly and individuals with diminished health. The proposed
medicine dispenser proves that has less time complexity and making the patient self-dependent on contrary
situations.
Keywords— IoT;Hearbeat Sensort; Servo motor; RFID; Pilbox
PAPER : EC851
Energy Efficient Smart Street Light System based on Pulse Width Modulation
Venkata Pullarao Matta, Ravi Sankar Miriyala, K V S H Gayatri Sarman, K V Subba Reddy M, M
V Pathi A, Venkateswara Rao Ch
[email protected]
Abstract— Across the country, majority of the electric power has been utilized for street lights. These
street lights are always ON throughout the night and even in day times due to inefficient street light
monitoring systems. An enormous rise in population and waning of natural resources has necessitates the
prerequisite of efficient street light system. Therefore, in this work an energy efficient smart street light
system based on PWM and Arduino has been proposed. In this work, the power efficiency has been
achieved through PWM by switching the brightness of the LED from 100% to 20% based on the vehicle
traffic. Based on the vehicle traffic per hour, the duty cycle has been adjusted which inherently changes the
width of the pulse. Initially, the lights are in OFF during day time regardless of vehicle movements(LDR
is high); during night time or cloudy environment the lights are turn ON based on the vehicle motion
detected by PIR. The brightness of the LED is 100% up to a specified time delay, thereafter the brightness
has been reduced to 20%. These changes in stats of LED (ON/OFF) have been controlled by PWM duty
cycle which is specified by analogWrite(). The performance measures have been evaluated for proposed
system by means of power consumption. The experimental results demonstrated that, the proposed system
archives a power save (delay of 3 seconds) of 95% with zero vehicle movement, 92% with 30 vehicles per
hour, 91% with 50 vehicles per hour and 87% with 100 vehicles per hour when compared to conventional
street light systems.
Keywords— Arduino; duty cycle; electricity; efficiency; light dependent resistor; power consumption;
passive infrared sensor; street lighgt system.
PAPER : EC854
IoT-based Intelligent Modeling of Security Monitoring System for Human Safety
Bhawna Khokher, Dhivya M., Devarasi Swetha, Anila D, Abhijit G Naik, Shwetha S
[email protected]
Abstract— Security has grown to be a top priority for women, kids, and even seniors from all walks of
life. Women are being attacked and molested, kids are being kidnapped, and elderly people are also dealing
with many issues like robberies. According to a recent WHO poll, 35% of women worldwide experience
some kind of abuse or physical assault. The number of victims is progressively rising. Here we introduce
a system which makes sure human protection whenever they sense danger. Numerous IoT (Internet of
Things) software systems are designed to support humans in difficult situations by alerting the contacts
they have registered. Our project resembles a smart device; with various sensors integrated. When any
185
ISBN: 979-8-3503-4821-7
person faces any kind of danger can press the button located on the band or locket, the various information
such as location, pulse rate, and the alert mechanism is employed. The system locates the user using the
Global Positioning System (GPS), notifies the emergency contact and police of the user's position through
SMS using the Global System for Mobile Communications (GSM), and notifies the registered contacts of
the victim's voicemail. This will aid in the reduction of crimes against humans.
Keywords— AURDINO, GPS, GSM, IoT, Human Security
PAPER : EC862
Artificial Intelligence based Recognition of Liver Melanoma using Sonogram Images
Dr.S.Gandhimathi alias Usha, Dr.S.Vasuki, Mrs.V.G.Janani, Ms.B.H.Lavanya, Ms.S.Seronica
[email protected]
Abstract— Liver cancer is one of the most malignant diseases and its diagnosis requires more
computational time. It can be minimized by applying a machine learning algorithm for the diagnosis of
cancer. The existing machine learning technique uses only the color-based methods to classify images
which are not efficient. So, it is proposed to use texture-based classification for diagnosis of tumor. The
input image is resized and pre-processed by Gaussian filters. The features are extracted by applying Gray
Level Co-occurrence Matrix (GLCM) and Local Binary pattern (LBP) in the preprocessed image. LBP is
an efficient texture operator which labels the pixels of an image by thresholding the neighborhood of each
pixel and considers the result as a binary number. The extracted features are classified by Multi-Support
Vector Machine (Multi SVM) and K-Nearest Neighbor (K-NN) algorithms. The advantage of combining
SVM with K-NN is that SVM measures a large number of values whereas K-NN accurately measures point
values. The results obtained from the proposed techniques achieved high precision, accuracy, sensitivity
and specificity than the existing method.
Keywords— Ultrasound imaging, multi-Support vector machine, kernel, GLCM, LBP
PAPER : EC871
Design And Optimization Of Microchannel For Neonatal
Ms.T. Archana,M.E., Karunya I, Krithika R, Subiksha V
[email protected]
Abstract— New development in technical progress on a worldwide platform is the development of
microsystems, micro - electromechanical, and similar technologies. The infusion pump that is typically
used to administer intravenous therapy to neonates will be replaced in this study with a microchannel.
Microchannels are defined as having a diameter of less than 1 millimeter, or between 1 and 99 micrometer.
Neonatal intravenous lines utilize microchannels to control the flows, velocity, and pressure of the fluids
and drugs. In certain types of emergencies, new borns demand nearly long, 24 hour medical assistance. In
such cases, they could not only deliver healthcare facilities. Microchannels can be used in place of such
infusion pumps to just make things simpler to observe and also care after new borns in homes even without
physician's supervision.
Keywords— Infusion pumps; Micro-channel; Intensive Care; Medication Error; Patient Safety; Neonatal,
Mems.
186
ISBN: 979-8-3503-4821-7
PAPER : EC874
Detection and Segmentation of Retinopathy Diseases using EAD-Net with Fundus Images
G.Sivapriya, V.Praveen, S.Saranya, R.Surya, S.Sweetha
Abstract— Diabetic retinopathy is one of the major concerns affecting most of the population. It causes
the injury in the blood vessels of the retina which is very sensitive to light and is located at the back of the
eye, causes this. In the early stages, it may did not cause any symptoms or it will be only minor vision
problems. When blood vessels are damaged, they can leak, causing dark spots to appear in our vision. The
Diabetic Retinopathy (DR) can be recognized by presence of Hard Exudate, Soft Exudates,
Microaneurysms and Haemorrhages. The most important aspect is accurate detection of diseases at an early
stage. Manually annotating these scratches is a significant task in clinical survey. This method is completely
based on the convolutional neural network and further that can be classified into three modules attention
module, encoder module and decoder module. The fundus images were normalised and augmented before
sent to the EAD-Net for pixel-wise label forecast and for Self-operating feature extraction. After pre-
processing, the image is sent into the EAD Net for training which is followed by testing of an image and
finally the segmentation of the image will be done. Optimizer is used here is Adam and categorical CE as
the loss function. This EAD-Net is the novel method for diagnosing different stages of DR. It produces
fitting results with an accuracy of 95 percentage when segmenting 4 different lesions. These active
segmentations have significant clinical implications in the monitoring and in the diagnosis of DR.
Keywords—Diabetic Retinopathy, Microaneurysms, Exudates, Haemorrh, insert
PAPER : EC876
A Review on Infectious and Cardiovascular Disease Classification Using Deep Learning

Dr.A.Sivasangari, G.V. Sai Kanth, CH. Sai Deekshith, G. Manohar, B. Arun Kumar, K. Vasavi
[email protected]
Abstract—The medical field is currently one of the most significant study topics because of the rapid
advancement of technology and data. Deep learning and machine learning are emerging techniques that
help with illness detection and prediction. Due to the fast growth of technology and data, the medical field
is today one of the most significant study topics. The authors propose that one approach to aid in illness
diagnosis and prediction is machine learning. Machine learning is used to identify whether a person has a
condition based on symptoms in the supplied data set. Machine learning methods are used to predict the
illness. A new model was developed using the required techniques for each disease in order to predict
whether or not a person had it. Even a user interface has been made available to patients so they may
manually input their symptoms and determine if they are ill. React, CSS, and Flask are the three most well-
known web technologies that are covered by this user interface.
Keywords—Deep Learning, CNN, Machine Learning, ML Techniques.
PAPER : EC893
Novel, Clock Gating Broadcasting Applications for Low-Power FPGA Architectures

Dr. Tata Jagannadha Swamy, Pranav Kumar
[email protected]
Abstract—In VLSI technology, clock gating strategies and designs are required for efficient power
utilization as well as for other design applications. In this connection, clock is one of the efficient tools for
Gating Broadcasting Applications for Low-Power FPGA Architectures. The proposed method mainly
187
ISBN: 979-8-3503-4821-7
explores the use of clock gating strategies to reduce the power utilization for streamed applications that
result from asyn- chronous dataflow architectures. Streaming applications include a wide variety of
computer methods from diverse fields as digital logic, digital content coding, encryption, etc. The
dynamically streaming nature of algorithms is taken into account in this research to present a set of
strategies that can produce power usage by selectively shutting off sections of the circuitry when they are
briefly dormant. These methods may be included into the synthesis phase of a high-level dataflow design
flow without regard to the semantics of the program being designed. At- scale implementations based on
field-programmable gate array platforms show that power may be reduced without impacting bandwidth
utilization, according to experimental studies.
Keywords—Clock Gating Strategies, Power FPGA Archi- tecture, Digital Content Coding, Encryption,
Data flow Design, Bandwidth Utilisation.
PAPER : EC902
HEALING AND PREVENTING TREES FROM BEETLES USINGPESTICIDES

Pradeepkumar G, Prabu M, Manimuthu Ayyannan, Pratheep V G, Revathi S, Neelam Sanjeev
Kumar
[email protected]
Abstract—The Indian way of life is extremely dependent on wildlife. Trees and their impact on products
are critical to the economy and societal advancement. These trees are also impacted by species such as
beetles during their growing season. Take, for example, banana trees, which are susceptible to beetle
infestation. This banana tree is vulnerable to beetle infestation at any time during its life. It is most common
during childhood. This can be harmful to the tree's health and shorten its life span in comparison to the
average. Pesticides for killing beetles are available on the market, but they must be applied manually to the
affected area each time. Some farmers may fail due to the difficulty of repeatedly climbing to apply
pesticide to a specific infected area; this method must be used for a specific period of time. Furthermore,
pesticides differ depending on the type of beetle, making it difficult for the farmer to select the appropriate
pesticide for the correct beetle. Cosmopolites sordidus is a 12 mm long insect that feeds on banana trees.
As a result, a syringe-like structured beetle killer was created. The syringe can be loaded with any pesticide
and then injected into the affected tree. Another goal is to reduce manpower and 24/7 monitoring.
Keywords—Agriculture, Banana tree,Cosmopolites sordidus beetle, Pesticides, 24/7 monitoring, Attacks
the growing period.
PAPER : EC908
Vehicle Detection and Classification using Deep Learning for Hazy Images
Krishna Dharavath, Mudigonda Ananya, B. Shashi Kanth, V. Vivek
[email protected]
Abstract—Haze is created as a result of air light and attenuation which reduces image contrast and air
lighting makes images whiter. An image can be made less white and have more contrast to dehaze it. For
advanced driver assistance systems to work, vehicle detection is a fundamental component. For which the
removal of haze or dehazing of the image is the most efficient solution to the problem of vehicle
recognition. One of the important dehazing techniques is called Dark Channel Prior (DCP), which relies
on observing the haze-free image's distinguishing features. This paper discusses a detailed examination of
the single picture haze reduction with the dark channel prior, proceeded by vehicle detection using the
GoogleNet deep learning method.
Keywords—Vehicle detection; Hazy images; Dark channel prior; GoogleNet; deep learning
188
ISBN: 979-8-3503-4821-7
PAPER : EC924
A Hierarchical Waste Segregation Using Smart Dustbin

K. Sony, K.Neeharika,K. Jahna Pravallika, M. Ajay Kiran, T. Akhil
Abstract—Technologies have always made it easier and better for humans to live. We are making progress
toward a healthier India, and we are currently implementing a revolutionary plan that will revolutionize the
waste management sector. Overflowing trash cans are a common sight in public places where adequate
waste disposal is not being done. Even spotless private areas couldn't utilize the resources at hand
effectively. Waste must be carefully managed, processed, and disposed of in order to minimize risks to
public safety and the environment. An enormous rise in the volume and type of solid waste is being
attributed to urbanization, continued economic growth, and industry. This method is a simple and quick
way to tell dry rubbish from wet garbage. It's designed to divide rubbish into dry and moist waste, which
may then be treated separately for the next operation. Continually monitoring the garbage can with
embedded technology and IoT to detect if it is fully or partially filled An LCD screen installed outside the
trash can displays a message if there is more waste than the containers can store as a result of wireless
sensors that detect this. We have many sensors that make the trash can smarter, so when there is motion in
front of it, the door automatically opens. The essential parts of the system of this gadget are the Arduino
UNO and ultrasonic sensors. A moisture sensor built inside a microprocessor and a conveyor belt was used
to create a smart trashcan for garbage separation in order to identify damp, dry, and recyclable material.
The integrated prototype produced adequate results to back up the tried-and-true method. With this
endeavor, the waste management system is taking the first steps toward creating a successful system that
will produce a sustainable environment for the present and a gorgeous, clean society in the next years.
Keywords—Segregation, Clean society, Metallic, Automatic, Garbage disposal, Microcontroller,
Sustainable environment
PAPER : EC943
Artificial intelligence for IoT-based Healthcare System
M vara siddardha reddy, R sai prasad, R sai jagan, Dr.M.Selvi,M.E.,Ph.D.
[email protected]
Abstract—Artificial intelligence (AI) and communication networks at 6G speeds are going to
fundamentally alter modern life. This will have repercussions for the medical industry. It is anticipated that
6G would eradicate the existing clinical hurdles. 6G is also a medical improvement. In light of this, our
proposed work on 6G healthcare is as follows: This article investigates the use of artificial intelligence (AI)
and the internet of things (IoT) inside the medical infrastructure, as well as the outcomes in a few clinical
fields.
Keywords—Health Care System, Artificial Intelligence Internet of Things, Deep learning and 6G.
PAPER : EC975
Age group determination using Logistic regression through a biometric approach
Srinithya Gunti, Shivani Ganti, Radam Kushaal
[email protected]
Abstract—Every person is authenticated using their natural physical or behavioralcharacteristics in the
process of biometricauthentication. When users enrolled in the models, many of them used soft biometric
data such their gender, age, height, and eye color. Years of work have gone into figuring out how old a
human face is roughly. Age assessment based on the eyes, particularly using the iris- pupil ratio, is a
relatively novel idea. The suggested approach attempts to predict age group using logistic regression using
189
ISBN: 979-8-3503-4821-7
this putative iris characteristic. Logistic Regression is a type of statistical model and it is used to determine
or predict the probability of an outcome and this is done by analyzing the relation between variables. After
segmenting iris images, the training and classification are derived from the iris-pupil ratio analysis.
Keywords—Logistic Regression, CASIA
PAPER : EC988
Advance IOT based air pollution detection and control system in vehicles
Ishita Deb, Swati Nigam, V Shruthi, Vinay Kumar KN, Krishna Chaithanya S, Varshitha B
[email protected]
Abstract—Esoteric automobiles are an important component of everyone's everyday routines. In some
circumstances and scenarios, the fast-paced metropolitan life necessarily requires the use of a vehicle.
Because of coin has multiple facets, each side has its own impact, among which is air pollution. Any vehicle
will diverge, but when the formalized value has been exceeded, issues occur. The principal reason of this
deviation stance ambiguity is totally inadequate combustion of the power delivered to the device, which
would be caused by poor vehicle support. These emigrations from your vehicle were unable be wholly
bypassed and although bearable. Considering advancements in semi-guided detectors for distinguishing
vibrantly colored feasts, this document describes the use of these semi-guided sensors in vehicular channels
to differentiate the focus of hazardous goods and describe this. That is the primary objective. When the
oxide layer position surpasses an ordinarily hard cap, the car will sound a buzzer to demonstrate that
perhaps the arrestment has indeed been surpassed, as well as the motor will be closed for a set timeframe
(its time allowed for the car driver to leave the car). Throughout this time, the Global positioning system
will begin to look for near the area aid kits. When the timekeeper expires, the machine's energy should be
disengaged, and the vehicle should be hauled away to a technician or the closest aid station. A small
regulator oversees and control mechanisms the syncing and prosecution of all commerce. When expanded
as a long mission, this article would then benefit the wider populace and assist in the decrease of pollution.
Keywords—Detectors, threshold level, GPS, and a micro – controller.
PAPER : EC989
Sensor Based Sewage Segmentation

Ishita Deb, Nayana G H, Ashritha Challa, Swathi A, Bandaru Kusuma Sagar, MD Afsar
[email protected]
Abstract—These days we are facing many problems because of garbage waste in this smart world .This is
because we are throwing our full house waste without segregating due to this manual help is not possible
for decomposing., if we segregate the waste into 3 possible ways like metal waste, wet and dry waste we
can decompose it in many natural ways and we reuse the waste by recycling process with the smart bins.
In order to segregate the house waste and industrial waste we are newly creating garbage segregation device
using sensors
Keywords—smart world, decomposing, smart bin, Sensors, recycling
PAPER : EC990
A Systematic Technique Used For Portable Telemedicine System
Monika Gupta, Swati Nigam, Yathika Evangelin, P Sumyukthaa, Rakshitha L
[email protected]
Abstract—With the increasing health issues in today’s world, healthcare technology has grown to be one
of the most popular studies nowadays. With the increasing demand for doctors and long queues to wait for
190
ISBN: 979-8-3503-4821-7
a chance to visit the doctor for the patient’s health to be monitored can be hectic and tiring depending on
the patient’s condition. With the progress of healthcare technology, the lifecycle of individuals has
magnificently stretched, yet lacks so much facilities and a ratio of 1:854(1 doctor for every 854 patients).
Still, individuals in the rural zone are still having a tough period to find specialized healthcare services due
to the distance and insufficient number of doctors. Subsequently, we are proposing an innovative technique
that dodges the need to find a doctor immediately by using a Portable Health Monitoring device, that uses
sensor technology and uses internet to communicate with your doctor and also prescribe the required
medication without any hustle. A portable patient monitoring device is one of the finest answers to
overcome this problem. This paper offers an Internet of Things (IoT) based real-time portable patient
monitoring device that is able to assure the veracity of the real-time.
Keywords—IoT, electrocardiography (ECG), Temperature Sensor, Heart Rate sensor, Pulse Rate Sensor,
Machine learning
PAPER : EC991
Comparative Study to determine Accuracy for Weather Prediction using Machine Learning
Swati Nigam, Monika Gupta, Akshaya Shrinivasan, Araveti Venkata Sai Uttej, Chaitra Kumari,
Disha P B
[email protected]
Abstract—As there are different Weather Prediction algorithms, this paper gives a comparative study to
determine accuracy for weather prediction on the basis of temperature, rainfall, humidity pressure etc. The
objective is to design an accurate weather prediction classification model.
Keywords—Machine learning, weather sensors, ESP8266, weather prediction, weather station
PAPER : EC992
Smart Use of Solar Radiation in Agriculture Purpose
Monika Gupta, Nithin Kumar NR, M Rajasekhar, Manika Sai Kiran, M.B. Rohith Reddy
[email protected]
Abstract—A promising renewable energy technique is solar energy. Due to the rotation of the Earth,
conventional stationary solar panels have a limited region of sun exposure. This issue is resolved by the
dual-axis solar tracker automated solar tracking technology. This project's goal is to address the energy
crisis, which is a significant challenge in the daily world and agriculture industry. Compared to traditional
energy producing techniques, the need for employing renewable energy sources, namely solar energy, has
grown. The microcontroller will decide how much rotation to perform depending on inputs it received from
an LDR close to the solar panel. The results show that, when compared to fixed panel systems, the
developed low-cost sun tracker increases output power production efficiency by 25-30%. And this
technique we can use it in agriculture needs where we require more energy for water pumps and other
purposes. So, it is sustainable and ecofriendly.
Keywords—Dual–Axis sun tracker, LDR, Photovoltaic cells, solar energy, solar panels
PAPER : EE171
A Review on the Analysis of 5G Technology and its Impact on Humans
R.DHANASEKAR, L.VIJAYARAJA, MIRTHULAA C S, NARENTHIRA SAI RAAM P P,
NAVEEN B, VIGNESH KUMAR C
[email protected]
Abstract—This paper is made to clarify the detailed information of 5G radiation. An overview on the
information on the consequences of 5G radiation on the human body and its architecture in 5G with related
to its security. There are numerous organizations set up to highlight the negative effects of 5G in a human's
191
ISBN: 979-8-3503-4821-7
day-to-day life. This acknowledges the public's understanding of 5G radiation and clarifies its benefits and
drawbacks. The 5G has already being used in many developed and developing countries. As the technology
develops the human body adopts with the development to restrain its presence in the living beings. The
assumption on the paper is given for a better simplification.
Keywords—Radiation, Health, Mobile phones, Data, Human body, frequency
PAPER : EE240
Design of mulitilevel inverter with its study on different configurations and applications
VIJAYARAJA L, DHANASEKAR R, S KESAV SANADHAN, VISHAL T, KANNAN V, RUPA
KESAVAN
[email protected]
Abstract—Electrical device that convert direct current to alternate are termed as inverter. It will be
interesting to present an inverter to produce a multilevel output that overcomes the conventional inverters.
This made the researchers to introduce new multilevel inverter that can be type of cascading or single unit
to produce higher output levels across the load. Here a detailed analysis is made to study the types of MLI
that can be made using less number of switches and other components in order to make it compact, so that
it will be placed in suitable applications. Study reveals that MLI produces higher level with fewer harmonic
that can be placed in electrical applications. Also an attempt is made to design a MLI using
MATLAB/Simulink with results.
Keywords—Diodes, multilevel inverters, PWM, Single-stage.
PAPER : EE486
DESIGN OF AN ENERGY EFFICIENT LOW VOLTAGE SINGLE PHASE IM
A. Selvanayakam, B. Hemananth, M. Iswarya, A. Elakya, N. Subhashini
[email protected]
Abstract—In order to eliminate high voltage human dangers, costly and difficult high-voltage isolation,
and requirements for electronic power devices, this study provides the design, analysis, and implementation
of the suggested single-phase induction motor. A conventional 220V 1-ph induction motor has been
changed in order to obtain the necessary 160V–180V, 1-ph motor. There have been theoretical and practical
comparisons between the traditional induction motor and the wound motor, as well as other design analyses.
Commercial components have been used in the implementation of this suggested design.
Keywords—Windings, Induction motors, Rotors design, Stator design, Air gaps
PAPER : EE487
GENETIC BASED FUZZY LOGIC CONTROL OF SEDC MOTOR
N. Aarthi, P. Anbarasu, D. Nagarajan, A. Sajitha Banu, M. Vinosh
[email protected]
Abstract—This thesis primarily aims at offering an efficient technique of speed control for the small,
independently excited SEDC motors utilised in a variety of applications, including industrial, commercial,
and medical. The major objective of this work is to suggest a practical approach for controlling the speed
of these weak motors. The natural optimization technique known as the genetic algorithm is employed in
the suggested way to enhance the speed-controlled operation of the SEDC motor. The goal of this thesis
work is to improve the values of several Performance parameters, such as rising time, time taken to settle,
time taken to fall, peak overshoot, and steady state error, in order to regulate the motor in an efficient
manner. The motor is operated using both the conventional PI controller and GA optimized controller
192
ISBN: 979-8-3503-4821-7
MATLAB version R2013a was used to generate the SIMUINK MODEL for both controller operations. In
terms of SEDC motor control, the proposed GA-optimized controller performs the best
Keywords—SEDC; rising time; settling time; optimization; performance.
PAPER : EE553
Implementation and Classification of Breast Cancer histopathological image Using Support Vector
Machine Learning
DEETCHITH KAISEREA, Dr.B.KIRUTHGA, Dr.R.NARMATHA BANU
[email protected]
Abstract—Deep learning (ML) is commonly used in the categorization of breast cancer (BC) pattern due
to its advantages in modelling important image features from complex BC information. In our proposed
system for autonomous prediction and diagnosis of BC using an ensemble of classifiers, different image
areas are defined by different tissue types that have incredibly variable looks. Tumor prognosis, image
categorization, and performance improvisation make up the three main contributions. Our approach is
evaluated on a publicly available dataset. The performance of the application is then evaluated using
performance improvisation and predictions. Our approach is evaluated on a publicly available dataset.
Lastly, the procedure's efficacy is evaluated using accuracy, sensitivities, and precision.
Keywords—Health Care System; deep learning; breast cancer ; image processing; SPV
PAPER : EE578
MODIFIED PI CONTROLLER FOR BLDC MOTOR IN ELECTRIC VEHICLE APPLICATIONS
Mr.G.Jayabaskaran, Dr.S.Suresh
[email protected]
Abstract—The recent dramatic growth in electric vehicles has brought renewed attention of their motors
and power electronics. Of the motors used for electric vehicles, the DC motor and the three phase induction
motor are well studied. Developing countries have created their own solutions for electric vehicles,
especially with the Brushless DC Motor (BLDC). The less investigated three phase BLDC motor presents
challenges for control of the transient and steady state response, overshoot, rise time, settling time, etc. A
lapse in the control can cause the system to become unstable and reduce the life of com ponents. In this
paper, a BLDC motor for electric vehicles is studied with the Modified Proportional Integral (MPI)
controller and logic gates for the drives. The design allows reduction 0.74 % (1 000 rpm) and the settling
time to 0.5 of overshoot to an estimated seconds. This study is expected to contribute to the applications of
the BLDC, especially for electric vehicles. The design, Analysis and Simulation of Current Iterative
Learning method is simulated using MATLAB SIMULINK R2022
Keywords—Brushless DC motor drives, Torque Ripple Reduction, Current control, Modified Proportional
Integral (MPI) Controller
PAPER : EE656
An Overview on Development & Prospects of Geothermal Power Plant in India
Namra Joshi, Dr. Achala Khandelwal
[email protected]
Abstract—With the increase in power demand, India is moving from conventional power generating
sources to non- conventional sources. The steady progress of economy and dearth of energy has led the
focus towards the geothermal energy. Geothermal energy is a convenient medium of power generation
having excellent potential in some regions of India. As per Geological Survey of India, around 340 sites
193
ISBN: 979-8-3503-4821-7
are identified to have geothermal potential. This article delivers a general framework on geothermal power
and highlights the fundamentals of geothermal power plants and their Indian Scenario.
Keywords—Renewable Energy, Geothermal Power; Indian Context
PAPER : EE673
REVIEW OF HYBRID WIND-SOLAR PV TECHNOLOGY IN THE GENERATION OF

ELECTRICITY
Jayakumar T, Ponmurugan P, Karthik B, Anandhan K C, Ragul S, Neelam Sanjeev Kumar,
Rajkumar D
[email protected]
Abstract—Achieving sustainability by utilizing alternative energy sources viable technological
possibilities for creating sustainable energy, the sun, biomass, wind, geothermal resources, hydropower,
and ocean resources are considered. Despite the fact that the total amount of energy produced by PV cells
and wind turbines is still far less than that of fossil fuels, their ability to generate electricity has significantly
expanded in recent years. This article provides an overview of the Solar-Wind hybrid power system, which
generates electricity by combining the Sun and Wind, two renewable energy sources. Microcontrollers are
widely used in the field of system management. We can maximize the utilization of those resources by
employing this strategy, which takes into account the distinct production processes of each resource.
Furthermore, it increases dependability and decreases reliance on any single input. This hybrid solar-wind
power generation system is suitable for both industrial and residential applications.
Keywords—Solar, Wind, Power Generation
PAPER : EE674
Portable Water Quality Detector
P.S.Manoharan, P.Dhanya, R.Sowmya, S.B.Sri Varuni, R.K.C.Jeykumar
[email protected]
Abstract—The Quality of water can be determined by water quality detector.TDS sensor gives the total
amount of solids dissolved in the water. Parts Per Million (PPM) is the unit of TDS. TDS sensor is
integrated to Arduino and the value can be displayed using LCD display. pH sensor is used to find the
concentration of hydrogen ions. If the concentration of hydrogen ions is high then the solution is acidic and
if the concentration of hydrogen ions is low then the solution is basic. pH sensor is also integrated to
Arduino. Then it compares the value of TDS and pH, displays the purpose of water. It displays both the
values and purpose of the water.
Keywords—TDS; pH; Water Quality
PAPER : EE921
Solar PV Fed Modular Multilevel Inverter with Fuzzy Logic Control for Marine Water Pumping
Applications
Sindhuja R, Kadipaku Chanduja Reddy, Gummalla Shamitha Reddy, Gurvindapalli Uday Kumar,
NaluguriVaishnavi Divya, Illuri Prem Kumar
[email protected]
Abstract—This paper presents the implementation of the intelligent techniques to drive the induction
motor used in the application for pumping water in ships. The eleven-level inverter that is being suggested
can regulate the speed of an IM drive that is powered by solar photovoltaics. With an induction motor (IM)
drive and MMI, the proposed control architecture for a marine water pumping system is investigated and
validated in this study. The inverter is evaluated using Proportional- Integral (PI) , Fuzzy Logic (FL) and
194
ISBN: 979-8-3503-4821-7
ANFIS logic driven controllers to improve its performance. A comparative analysis will be conducted
towards increasing of robustness of the inverter in terms of Total Harmonic Distortion(THD) , peak over
shoot andcontroller settling time. The design and combination of IM Drive, MMI, and controllers
exclusively for ocean pumping stations applications make the proposed control method innovative.
Keywords—Induction Motor driver circuits, Intelligent Controllers, Modular Multilevel Inverter, Total
Harmonic Distortion
PAPER : EE973
A New Fault Tolerant Five-Level Multilevel Inverter for Critical Loads
B. Hemanth Kumar, Vinod Bakka, B. Suresh, Chandrika V S, K. Janardhan, R. Seshu Kumar
[email protected]
Abstract—Abstract - This work offers a fault tolerant (FT) multilevel inverter that may operate
continuously even when an open switch fault (OSF) occurs in a single or several power switches. The
number of power semiconductor devices used in dc-ac converters has increased in order to obtain high-
quality voltage waveforms, making them prone to failure. As a result, one of the most significant issues in
the use of multilayer inverters (MLIs) in many industrial applications is reliability. Simultaneously, reduced
component multilevel inverters (RC-MLIs) can obtain the highest resolution in output voltage waveform
by sacrificing critical aspects (such as fault tolerance ability, charge balancing control, and so on), but it
must be kept. In terms of switch level failures, redundant switching states have been prioritized for effective
fault tolerant operation of MLIs. In the event of an open circuit fault on single or many switches, a study is
being conducted to design an unique fault-tolerant MLI. As a result, the suggested FT-MLI structure can
meet the operating requirements for the time being and is said to have fewer components. To generate the
desired output voltage waveforms, a valid fault tolerant switching approach is used in conjunction with
sufficient fault clearance.
Keywords—Multilevel inverter, Fault-tolerance, Switch fault, Reliability.
PAPER : EE977
PERFORMANCE INVESTIGATION OF RESONANT CONVERTER WITH AND WITHOUT
CONTROLLER
Dr.N.Madhanakkumar, Dr.Arulkumar Subramanian, Dr.M.Vijayaragavan, V.Sriramkumar
Abstract—The surviving on present literature focusing the design of resonant converter to work on various
drives. In this literature, the PI controller is subjected to perform co-ordinates with resonant converter with
DC input voltage of 55V for 80ohms nominal load. The designed resonant converter is exposed to both
line and load side disturbances. The time domain specification (TDS) of proposed resonant converter is
observed. The positive and effective feedback is come from PI controller combining with proposed resonant
converter than the common designed converter. The time domain provisions like Delay time, Rise time,
Peak time, Peak Overshoot and Settling time of PI controller co-ordinates with proposed resonant converter
is basically effective than resonant converter working alone.
Keywords—Resonant Converter, PI Controller
PAPER : EE978
CONSTANCY INVESTIGATION OF FUZZY CONTROLLER OVER PROPORTIONAL
INTEGRAL CONTROLLER FOR RESONANT CONVERTER
Dr.N.Madhanakkumar, Dr.Arulkumar Subramanian, Dr.M.Vijayaragavan, V.Sriramkumar
Abstract—In this article mainly aimed to reduce the disturbance raised in the line or load side of resonant
converter. We have to achieve effective output from the resonant converter among any disturbance
195
ISBN: 979-8-3503-4821-7
condition. So that the previous literature survey the resonant converter is subjected to line and load side
disturbances. In the survey told that the resonant converter works along with the help of controllers will
give the better output when compared with the resonant converter working alone. Here the designed
resonant converter is manually subjected to line and load disturbances with input voltage of 55V supplying
for 80ohms load which the converter is integrating with Proportional Integral (PI) and Fuzzy logic
controller. The designed converter’s disturbance output voltages and currents waveforms of PI and Fuzzy
controllers are compared by time domain specifications like peak time, raise time, settling time and peak
overshoot are measured by using output waveforms and then we have to conclude that the recital working
of Fuzzy controller is more effective than the Proportional Integral (PI) controller.
Keywords—Resonant Converter, Proportional Integral (PI) Controller, Fuzzy logic Controller, Time
Domain Specifications (TDS), Peak Overshoot, Raise Time, Peak Time, Settling Time.
196

ICCCI Conference Proceedings

Uploaded by

Copyright:

Available Formats

ICCCI Conference Proceedings

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ICCCI Conference Proceedings

Uploaded by

Copyright:

Available Formats

13 Edition

2023 International Conference

Machine Learning &

Embedded Systems Industrial

Approved by AICTE, New Delhi Affiliated to Anna University, Chennai

2023 International Conference on Computer Communication and Informatics

January 23-25, 2023, Coimbatore, India

2023 International Conference on Computer Communication and Informatics

Copyright and Reprint Permission

Message from Chairman

Another important aspect is the interdepartmental research. In today’s world of internet of

My best wishes to the conference.

Message from Principal

Sri Shakthi Institute of Engineering and Technology.

Scope of ICCCI 2023

The 2023 International Conference on Computer Communication and Informatics (ICCCI

• Unique Opportunity to Partner for Maximum Visibility & Community Engagement

2023 International Conference on Computer Communication and Informatics

Date Slot Time Speakers

Conference Inaugural Session

Xavier Fernando, PhD, P. Eng.,

23-Jan-2022 FN 09 .00 am – 10.00 am Dr Nalinda Somasiri, BEng (Hons), MSc, PhD

Dr.Ramesh Kumar Ayyasamy

Conference Valedictory Session

Technical Program Chair

INTERNATIONAL ADVISORY COMMITTEE

Prof. Dr. Manuel Cedillo-Hernandez, Universidad Nacional Autonoma de Mexico

NATIONAL ADVISORY COMMITTEE

Dr. K. Porkumaran, Chairman, IEEE Madras Section

S.No NAME DESIGNATION COLLEGE NAME

3 J.Bhuvana Assistant Professor King Khalid University

4 Ramu Avanashiappan Assistant Professor

6 Chandrasekharan Nataraj Senior Lecturer

7 Bela Genge Assistant Lecturer Petru Maior University of Tg. Mures

8 Xiao-Zhi Gao Professor

9 Seyed M. Buhari Associate Professor

10 Rosula Reyes Associate Professor

11 P.D.D.Dominic Associate professor

12 Wael Mohamed Khedr Assistant Professor

13 Ahmed Hassan Mohammed Hassan Assitance Professor

14 Hamid Ali Abed AL-Asadi Professor and Head Basra University

15 Yong Li Associate Professor

18 Selvakumar Samuel Senior Lecturer

23 Gwanggil Jeon Assistant Professor

24 Anand paul Associate Professor

25 Karen Cheung Professor The University of British Columbia

26 Richi Nayak Associate Professor

27 Nakhoon back Professor

28 Tor Aamodt Professor The University of British Columbia

29 Sigurd Meldal Professor

30 Tilahun Mamuye Gidey Technical Assistant

31 Xavier Fernando Professor

32 Alaa Hussein Al Hamam Assistant Professor

33 Hamid R Arabia Professor

34 Ahmed F.Zobaa Senior Lecturer

35 Dr. Arthur James Swart Associate Professor Central University of Technology

36 Dr. Santolo Meo Professor FEDERICO II University

37 Dr. Ismail Musirin Professor Universiti Teknologi Mara

41 Yuen , Henry Assistant Professor columbia university