Papers by Dr. K. Anitha Sheela
Implementation of ECG Based Person Authentication System Using NUFB Features and GMM-UBM
This paper goals at studying the place and possible contribution of "Internet of Things" (IoT) in... more This paper goals at studying the place and possible contribution of "Internet of Things" (IoT) in the context of the EU's ambitious climate and energy targets for 2020. Using qualitative procedure, we are mainly concentrating on Demand Side Management (DSM) as an effective method in balancing the load of Electrical Distribution Networks. The role of IoT in DSM is to enable and enhance electrical energy peak demand reduction and its maximum uniform time-distribution achieved through society's ecoeducation. Using computational tools such as Data Mining and Recommender System we can achieve results at the level of electrical energy distribution network reflected in peak reduction and its uniform time distribution.
Recently, WSNs have drawn a lot of attention due to their broad applications in both military and... more Recently, WSNs have drawn a lot of attention due to their broad applications in both military and civilian domains. Data security is essential for success of WSN applications, exclusively for those mission-critical applications working in unattended and even hostile environments which may be exposed to several attacks. This inspired the research on Data security for WSNs. Attacks due to node compromise include Denial of service (DoS) attacks such as selective forwarding attacks and report disruption attacks. Nearby many techniques have been proposed in the literature for data security. Hop-hop security works well when assuming a uniform wireless communication pattern and this security designs provides only hop-hop security. Node to sink communication is the dominant communication pattern in WSNs and hop-hop security design is not sufficient as it is exposed to several attacks due to node compromise. Location aware end-end data security (LEDS) provides end-end security.
Air quality has been largely affected by industrial activities, which have caused many health iss... more Air quality has been largely affected by industrial activities, which have caused many health issues among people. Air pollutants levels can be measured using gas sensors. Internet of Things (IoT) technology can be used to remotely detect pollution. IOT devices can be used to control basic functions from anywhere around the world through the internet. The data gathered by such a system can be transmitted instantly to a web-based application to enable monitoring real time data and allow immediate risk management. The aim is to build a system which can be used to monitor the parameters in different environments. Here, we describe an entire Internet of Things (IoT) system that collects real-time data in specific locations. This real-time data collected is compared with a predetermined threshold. This data is sent to the concerned organization notifying them about the values exceeding the threshold if any and take necessary actions if needed.
A novel wavelet based approach for near lossless image compression using modified duplicate free run length coding
Applied Computer Science, 2014
In this paper we are presenting a three-stage near lossless image compression scheme. It belongs ... more In this paper we are presenting a three-stage near lossless image compression scheme. It belongs to the class of lossless coding which consists of wavelet based decomposition fol- lowed by modified duplicate free run-length coding. We go for the selection of optimum bit rate to guarantee minimum MSE (mean square error), high PSNR (peak signal to noise ra- tio) and also ensure that time required for computation is very less unlike other compres- sion schemes. Hence we propose 'A wavelet based novel approach for near lossless image compression'. Which is very much useful for real time applications and is also compared with EZW, SPIHT, SOFM and the proposed method is out performed.
Wavelet based Effective Color Image Compression using Neural Networks and Modified RLC
Digital Image Processing, 2012
Image compression is a technique of reducing the size of image by eliminating data redundancy. It... more Image compression is a technique of reducing the size of image by eliminating data redundancy. It helps in reducing the amount of memory required to store an image and the time required to transmit the image over long distance. Earlier image compression is performed by using wavelet and neural network. This paper proposes a method for image compression that uses wavelet and Multilayer Feed forward neural network (MLFFN) with Error Back Propagation algorithm (EBPA), which is used to train multi layer feed forward neural network with an excellent input and output mapping. This algorithm is used for LL2 component and Modified Run Length Coding (RLC) to LH2, HL2 components with hard threshold to discard insufficient coefficients. Performance of proposed image compression method is evaluated using Compression Ratio (CR), Peak Signal to Noise Ratio (PSNR), Mean Square Error (MSE). These estimation parameters were found to be greater when compared to image compression methods SOFM, EZW, SPIHT.
Hyper Spectral Image compression using Higher Order Orthogonal Iteration Tucker decomposition
2022 IEEE 19th India Council International Conference (INDICON), Nov 24, 2022
DOAJ (DOAJ: Directory of Open Access Journals), Oct 1, 2019
International journal of circuits, systems and signal processing, Oct 7, 2022
Tensor decomposition methods have beenrecently identified as an effective approach for compressin... more Tensor decomposition methods have beenrecently identified as an effective approach for compressing high-dimensional data. Tensors have a wide range of applications in numerical linear algebra, chemo metrics, data mining, signal processing, statics, and data mining and machine learning. Due to the huge amount of information that the hyper spectral images carry, they require more memory to store, process and send. We need to compress the hyper spectral images in order to reduce storage and processing costs. Tensor decomposition techniques can be used to compress the hyper spectral data. The primary objective of this work is to utilize tensor decomposition methods to compress the hyper spectral images. This paper explores three types of tensor decompositions: Tucker Decomposition (TD_ALS), CANDECOMP/PARAFAC (CP) and Tucker_HOSVD (Higher order singular value Decomposition) and comparison of these methods experimented on two real hyper spectral images: the Salinas image (512 x 217 x 224) and Indian Pines corrected (145 x 145 x 200). The PSNR and SSIM are used to evaluate how well these techniques work. When compared to the iterative approximation methods employed in the CP and Tucker_ALS methods, the Tucker_HOSVD method decomposes the hyper spectral image into core and component matrices more quickly. According to experimental analysis, Tucker HOSVD's reconstruction of the image preserves image quality while having a higher compression ratio than the other two techniques.
Real-Time Emotional Analysis from A Live Webcam Using Deep Learning
2022 3rd International Conference for Emerging Technology (INCET), May 27, 2022
Emotion Recognition from Speech Biometric System Using Machine Learning Algorithms
Lecture notes in electrical engineering, 2021
Development of a Speech to Indian Sign Language Translator
Lecture Notes in Networks and Systems
Hyper Spectral Image compression using Higher Order Orthogonal Iteration Tucker decomposition
2022 IEEE 19th India Council International Conference (INDICON)
Hybrid Compression Method for Hyper Spectral Images using 3-Way SVD Tensor Decomposition and Discrete Wavelet Transform
2021 IEEE International Conference on Intelligent Systems, Smart and Green Technologies (ICISSGT), 2021
Hyper Spectral Image compression has attracted a lot of concentration in recent years due to mass... more Hyper Spectral Image compression has attracted a lot of concentration in recent years due to massive data volumes. Thousands of gigabytes make up hyper spectral datasets. The amount of important information included in these hyper spectral images is quite limited due to redundancy. The collected HSI data sets are dominated by spatial and spectral correlations. Many transform coding approaches are available for compressing hyper spectral images by reducing spatial dimensions. These algorithms were originally designed for grayscale or color images and have been extended to hyper spectral images. Tensor decomposition methods are developed to minimize the size of an N-dimensional function. The spatial and spectral dimensions of a hyper spectral image can be reduced by modeling it as a 3rd order tensor. In this proposed hybrid method, a Discrete Wavelet Transform is used to process the hyper spectral image, which is then decomposed into Core Tensor and Component Matrices using the 3 Way Singular Value Decomposition Tucker Decomposition method. Four datasets are used to evaluate the proposed method and observed SSIM, PCC, and PSNR performance factors are good with high compression ratio for datasets over the existing techniques.
ICTACT Journal on Microelectronics, 2019
Extraction and implementation of nonlinear model in computer aided design (CAD) tools requires mu... more Extraction and implementation of nonlinear model in computer aided design (CAD) tools requires multiple steps. In this paper we present framework for nonlinear parameters extraction and implementation of model in CAD tools. Methodology selection at each step is based on the requirement according to the LN G7A in house fabricated MESFETs device and its application. To validate the model, simulated sparameters from model are compared with measured data and percentage of error is below 10.3% from 500MHz to 20GHz. Measured and simulated drain to source current (Ids) is compared to validate the dc characteristics of the device. To evaluate the convergence of model in CAD tools, 5 to 5.5 GHz medium power amplifier is simulated which show that the model has not convergence issue during circuit simulation in Advanced Design System (ADS).
Emotion Recognition from Facial Biometric System Using Deep Convolution Neural Network (D-CNN)
Lecture notes in mechanical engineering, 2021
Advanced machine learning discriminant analysis models for face retrieval system
2018 2nd International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC)I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC), 2018 2nd International Conference on, 2018
advanced machine learning discriminant analysis models for face retrieval system has been impleme... more advanced machine learning discriminant analysis models for face retrieval system has been implemented in this paper. The image can be portioned into different subband system to extract the features. To reduce the dimensionality to extract a few features. So, it takes more computational time to extract the features. The proposed method can be implemented in one of the most popular methods is dimensionality reduction. Discriminate analysis models are widely applied on linear and non-linear expressions to extract the feature basis on scatter matrix. Different preserving projection models are used for extract the features from local and non-local matrices. In this paper, we proposed the advanced machine learning discriminate analysis models based on discriminant analysis and preserving projection based on the ORL, YALE database. This paper gives improved results regarding of face recognition parameters as compared to existing models.
HSDPA is a new feature which is introduced in Release-5 specifications of the 3GPP WCDMA/UTRA sta... more HSDPA is a new feature which is introduced in Release-5 specifications of the 3GPP WCDMA/UTRA standard to realize higher speed data rate together with lower round-trip times. Moreover, the HSDPA concept offers outstanding improvement of packet throughput and also significantly reduces the packet call transfer delay as compared to Release -99 DSCH. Till now the HSDPA system uses turbo coding which is the best coding technique to achieve the Shannon limit. However, the main drawbacks of turbo coding are high decoding complexity and high latency which makes it unsuitable for some applications like satellite communications, since the transmission distance itself introduces latency due to limited speed of light. Hence in this paper it is proposed to use LDPC coding in place of Turbo coding for HSDPA system which decreases the latency and decoding complexity. But LDPC coding increases the Encoding complexity. Though the complexity of transmitter increases at NodeB, the End user is at an a...
International Journal of Computer Applications, 2017
The person identification is an active area in research fields. many person identification techni... more The person identification is an active area in research fields. many person identification techniques have been proposed in literature both in time domain and transformed domain. An improved various transformed domain techniques are proposed in this paper. This paper work also demonstrates the task of identifying the person with the various segments of ECG signals, and also investigates which segments of ECG signals has more person specific information by using transformational methods. A transformed domain technique includes discrete fourier transform, discrete cosine transform and discrete wavelet transform. An experimental results on ECG signals using transformed domain techniques demonstrates that the improvement of proposed techniques over those of time domain techniques.
Uploads
Papers by Dr. K. Anitha Sheela