Presentation 2005-01-28
Lecture Video Sammary using Self-Organizing Map
Minoru NAKAYAMA, Takahiro YONEKAWA, Yasutaka SHIMIZU,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This report describes feasibility that scene thumbnail images of lecture video can be summarized using Self-Organizing Map (SOM). To extract a feature of scene thumbnail image from a lecture video, the discrete cosine transformation (DCT) was conducted for 15 segments of a image, then 3 major components from each segment were obtained. Thumbnail images were placed two-dimensionaly on the trained SOM according to each feature of the image. As results, similar images were classified on the SOM such as lecturer talking scene, presentation patterns and printed material images. Labels of presenting time provided the images, the time sequence of images also summarized on the SOM. The hybrid-type SOM which was trained with both image and term appearance features was also examined the performance of scence classfication.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Lecture video material / Video summary / Self-Organizing Map(SOM) / Indexing
Paper # ET2004-96
Date of Issue

Conference Information
Committee ET
Conference Date 2005/1/21(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Educational Technology (ET)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Lecture Video Sammary using Self-Organizing Map
Sub Title (in English)
Keyword(1) Lecture video material
Keyword(2) Video summary
Keyword(3) Self-Organizing Map(SOM)
Keyword(4) Indexing
1st Author's Name Minoru NAKAYAMA
1st Author's Affiliation CRADLE, Tokyo Institute of Technology()
2nd Author's Name Takahiro YONEKAWA
2nd Author's Affiliation HUB Networks, Inc.
3rd Author's Name Yasutaka SHIMIZU
3rd Author's Affiliation National Institute of Multimedia Education
Date 2005-01-28
Paper # ET2004-96
Volume (vol) vol.104
Number (no) 643
Page pp.pp.-
#Pages 6
Date of Issue