Presentation 2012-02-09
Subject Adaptation and Adaptive Training for Gait-based Person Identification
RASYID Muhammad AQMAR, Koichi SHINODA, Sadaoki FURUI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) The performance of gait-based person identification using statistical methods such as hidden Markov models (HMMs) often degrades when the amount of training data for each subject is insufficient. In our previous study, we proposed a gait-based person identification method using HMMs where we trained subject-dependent model from the scratch using each subject's training data. In this paper, we propose a model adaptation scheme, where the data from the other subjects are effectively utilized in the model training. We further improve the adaptation performance using subject adaptive training by effectively excluding the inter-subject variability from the subject-independent model. The proposed method improved the identification performance even when the amount of data was extremely small.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) CHLAC / FDA / PCA / MLLR / CMLLR / SAT
Paper # PRMU2011-199,SP2011-114
Date of Issue

Conference Information
Committee SP
Conference Date 2012/2/2(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 Speech (SP)
Language ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Subject Adaptation and Adaptive Training for Gait-based Person Identification
Sub Title (in English)
Keyword(1) CHLAC
Keyword(2) FDA
Keyword(3) PCA
Keyword(4) MLLR
Keyword(5) CMLLR
Keyword(6) SAT
1st Author's Name RASYID Muhammad AQMAR
1st Author's Affiliation Department of Computer Science, Graduate School of Information Science and Engineering, Tokyo Institute of Technology.()
2nd Author's Name Koichi SHINODA
2nd Author's Affiliation Department of Computer Science, Graduate School of Information Science and Engineering, Tokyo Institute of Technology.
3rd Author's Name Sadaoki FURUI
3rd Author's Affiliation Department of Computer Science, Graduate School of Information Science and Engineering, Tokyo Institute of Technology.
Date 2012-02-09
Paper # PRMU2011-199,SP2011-114
Volume (vol) vol.111
Number (no) 431
Page pp.pp.-
#Pages 6
Date of Issue