Scribd Logo
Upload

Welcome to Scribd!

  • 27 views

    Video:: Query-Focused Video Summarization: Dataset, Evaluation, and A Memory Network Based Approach

    Uploaded by

    Mark Tally
    This document discusses video summarization techniques. It introduces video summarization as allowing users to quickly browse lengthy videos by capturing the essence and removing redundant information. It describes applications as consumer video, browsing recorded content, and viewing interesting parts quickly. However, summarization also results in loss of information and is computationally expensive to evaluate. Two papers are described that propose neural network approaches for query-focused and query-conditioned video summarization and evaluation metrics. Both papers use datasets to train and evaluate their methods.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd

    Video:: Query-Focused Video Summarization: Dataset, Evaluation, and A Memory Network Based Approach

    Uploaded by

    Mark Tally
    27 views2 pages
    This document discusses video summarization techniques. It introduces video summarization as allowing users to quickly browse lengthy videos by capturing the essence and removing redundant information. It describes applications as consumer video, browsing recorded content, and viewing interesting parts quickly. However, summarization also results in loss of information and is computationally expensive to evaluate. Two papers are described that propose neural network approaches for query-focused and query-conditioned video summarization and evaluation metrics. Both papers use datasets to train and evaluate their methods.

    Original Description:

    Original Title

    Introduction

    Copyright

    © © All Rights Reserved

    Available Formats

    DOCX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document discusses video summarization techniques. It introduces video summarization as allowing users to quickly browse lengthy videos by capturing the essence and removing redundant information. It describes applications as consumer video, browsing recorded content, and viewing interesting parts quickly. However, summarization also results in loss of information and is computationally expensive to evaluate. Two papers are described that propose neural network approaches for query-focused and query-conditioned video summarization and evaluation metrics. Both papers use datasets to train and evaluate their methods.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    27 views2 pages

    Video:: Query-Focused Video Summarization: Dataset, Evaluation, and A Memory Network Based Approach

    Uploaded by

    Mark Tally
    This document discusses video summarization techniques. It introduces video summarization as allowing users to quickly browse lengthy videos by capturing the essence and removing redundant information. It describes applications as consumer video, browsing recorded content, and viewing interesting parts quickly. However, summarization also results in loss of information and is computationally expensive to evaluate. Two papers are described that propose neural network approaches for query-focused and query-conditioned video summarization and evaluation metrics. Both papers use datasets to train and evaluate their methods.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    of 2

    Introduction:

    Video:

     Video data is great asset for information extraction and knowledge discovery.
     Due to its size and variability, it is extremely hard for users to monitor.

    Video Summarization:

     Intelligent video summarization algorithms allow us to quickly browse a lengthy video by


    capturing the essence and removing redundant information.

    Applications:
    The application of video summarization can be divided into three main categories:

    1) Consumer Video Applications


    2) Browsing the recorded content
    3) View the interesting parts quickly

    Limitations:

     Loss of information
     Computationally expensive
     Evaluate the performance of a video summarizer
     No single video summarizer fits all users

    Related work:

    Paper 1:
    Query-Focused Video Summarization: Dataset, Evaluation, and A Memory Network Based
    Approach.

    Author introduces user preferences in the form of text queries propose a memory network
    parameterized sequential determinantal point process. Author also contend that a good evaluation
    metric for video summarization should Focus on the semantic information to collect dense per-
    video-shot concept annotations and new evaluation method.

    Objective Proposed Method Dataset Strength Limitation


    Main obstacles to Authors propose 1. UTEgocentric 1)Introduces user 1)Collecting
    the research on a memory net- (UTE) dataset preferences in dense per- video-
    video work the form of text shot concept
    summarization parameterized queries 2) Author annotations. 
    are the user sequential collect dense
    subjectivity-- determinantal per-video- shot
    users have point process to concept
    various attend the user annotations.
    preferences over query onto
    the summaries. different video
    frames and
    shots.
    Paper 2:
    Query-Conditioned Three-Player Adversarial Network for Video Summarization BMBC 2018

    Video summarization plays an important role in video understanding by selecting key frames/shots.
    Traditionally, it aims to find the most representative and diverse contents in a video as short
    summaries. In this paper, Author propose a query-conditioned three-player generative adversarial
    network to tackle this challenge. The generates or learns the joint representation of the user query
    and the video content, and the discriminator takes three pairs of query-conditioned summaries as
    the input to discriminate the real summary from a generated and a random one.

    Objective Proposed Method Dataset Strength Limitation


    Main aims to find Authors propose , 1. UTEgocentric Results are more 1)Do not
    the most a query- (UTE) dataset accurate based randomly
    representative conditioned on user query. generated
    and diverse three-player summary. 
    contents in a generative
    video as short adversarial
    summaries. network to tackle
    this challenge.
    The generator
    learns the joint
    representation of
    the user query
    and the video
    content.

    You might also like