Presentation 2011-02-17
A Multi-modal, Multi-frame Approach for Semantic Indexing in TRECVID
Nakamasa INOUE, Yusuke KAMISHIMA, Koichi SHINODA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We propose a multi-modal, multi-frame approach for semantic indexing in the TRECVID 2010 workshop. The goal of the semantic indexing is to develop a method for indexing many concepts that will be useful for video search. In this study, we implement a simple and accurate system by using Gaussian Mixture Model (GMM) supervectors with SIFT and MFCC features. The SIFT features are extracted not only from key-frames but also from many image frames in a shot in order to get the most out of multi-frame information. Our best result on the TRECVID 2010 dataset was 7.36% in terms of Mean Inferred Average Precision.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Semantic Indexing / Multi-modal / Multi-frame / GMM Supervector
Paper # PRMU2010-212
Date of Issue

Conference Information
Committee PRMU
Conference Date 2011/2/10(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 Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A Multi-modal, Multi-frame Approach for Semantic Indexing in TRECVID
Sub Title (in English)
Keyword(1) Semantic Indexing
Keyword(2) Multi-modal
Keyword(3) Multi-frame
Keyword(4) GMM Supervector
1st Author's Name Nakamasa INOUE
1st Author's Affiliation Department of Computer Sience, Tokyo Institute of Technology()
2nd Author's Name Yusuke KAMISHIMA
2nd Author's Affiliation Department of Computer Sience, Tokyo Institute of Technology
3rd Author's Name Koichi SHINODA
3rd Author's Affiliation Department of Computer Sience, Tokyo Institute of Technology
Date 2011-02-17
Paper # PRMU2010-212
Volume (vol) vol.110
Number (no) 414
Page pp.pp.-
#Pages 6
Date of Issue