A Study on Video Scene Retrieval Based on Online Video Annotation

PDF
Tomoki MASUDA
Graduate School of Information Science, Nagoya University

abstruction

1 Introduction

2 Video Annotation

2.1 Offline Video Annotation

2.1.1 about Offline Video Annotation

2.1.2 VAE(Video Annotation Editor)

Fugure1:

2.2 Online Video Annotation

2.2.1 about Online Video Annotation

2.2.2 Synvie

Fugure2:

Fugure3:

3 Tagging to Video Scene (create Scene Tag)

3.1 Tagging by Using Annotation Tool

Fugure4:

3.2 Automatic Extraction of Scene Tag from Online Video Annotation

3.3 Scene Tag Extraction Using Tag Selecting System

Fugure5:

Fugure6:

3.4 Experiment of Selecting Scene Tag

3.4.1 About

3.4.2 Result

3.4.3 Consideration

4 Video Scene Retrieval

4.1 Video Scene Retrieval System Using Scene Tag

4.1.1 Interface

Fugure7:

Fugure8:

Fugure9:

Fugure10:

Fugure11:

Fugure12:

4.1.2 Retrieval Algorithm

4.2 Experiment of Video Scene Retrieval

4.2.1 About

Fugure13:

4.2.2 Result

4.2.3 Consideration

5 Related Researches

5.1 Related on Online Video Annotation

5.1.1

5.1.2 View Generation of TV Content using TV Viewer's Viewpoint based on Live Chat on the Web

5.1.3 Development of Simple Movie Editing Application over the Network

5.1.4

5.1.5 Internet Street TV that vitalizes communication

5.2 Related on Offline Video Annotation

5.2.1 Informedia project

5.2.2 Web-Assisted Annotation

5.2.3 Tools for Video Annotation

6 Summary and Next Challenge

6.1 Summary

6.2 Next Challenge

6.2.1 Search of Non-Tagging Scene

6.2.2 Improvement about creating Scene Tag

6.2.3 Development of System for Acquiring more Annotation

6.2.4 New Application for Video

thanks