Structural Analysis of Instruction Utterances Using Linguistic and Visual Information

Tomohide Shibata*, Masato Tachiki, Daisuke Kawahara, Masashi Okamoto, Sadao Kurohashi, Toyoaki Nishida

*この研究の対応する著者

研究成果: Chapter

3 被引用数 (Scopus)

抄録

In realizing video retrieval system, the crucial point is how to provide an effective access method of video contents. This paper focuses on Japanese cooking instruction utterances and describes a method of analyzing structure of them, which leads to a summary of video. We detect a hierarchical structure of video contents by using linguistic and visual information. We found that the integration of visual information can improve the detection of task units better than using linguistic information alone.

本文言語English
ホスト出版物のタイトルLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
編集者Mircea Gh. Negoita, Robert J. Howlett, Lakhmi C. Jain
出版社Springer Verlag
ページ393-400
ページ数8
ISBN(印刷版)9783540301325
DOI
出版ステータスPublished - 2004
外部発表はい

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
3213
ISSN(印刷版)0302-9743
ISSN(電子版)1611-3349

ASJC Scopus subject areas

  • 理論的コンピュータサイエンス
  • コンピュータ サイエンス(全般)

フィンガープリント

「Structural Analysis of Instruction Utterances Using Linguistic and Visual Information」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル