Presentation 2005-10-28
Integration of Classifier for Gender and Age classification
Teruhide HAYASHIDA, Kazuya UEKI, Tetsunori KOBAYASHI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We propose a technique of integrating different types of classifiers for gender and age classification by using face images. We recognize gender and age with support vector machine(SVM) and Gaussian mixture model(GMM). We normalize the output value from each classifiers, and put it on the classifier again. Two normalization techniques are tested : one is posterior probability based one and the other is variance based one. Our experimental results show that our proposed technique improves classification rate compared to the classifier alone. There is a problem that time increases by integration of the classifier. Thus, we used two-dimensional principal component analysis(2DPCA) over convenional principal component analysis(PCA) for feature extraction. As a result, feature extraction time improved by about 50% and we achieived real-time performance on gender and age classification.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Gender and Age-Group Estimation / 2DPCA / GMM / Support Vector Machine
Paper # PRMU2005-96
Date of Issue

Conference Information
Committee PRMU
Conference Date 2005/10/21(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 Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Integration of Classifier for Gender and Age classification
Sub Title (in English)
Keyword(1) Gender and Age-Group Estimation
Keyword(2) 2DPCA
Keyword(3) GMM
Keyword(4) Support Vector Machine
1st Author's Name Teruhide HAYASHIDA
1st Author's Affiliation Science and Engineering, Waseda University()
2nd Author's Name Kazuya UEKI
2nd Author's Affiliation Science and Engineering, Waseda University:NEC Soft, Ltd.
3rd Author's Name Tetsunori KOBAYASHI
3rd Author's Affiliation Science and Engineering, Waseda University
Date 2005-10-28
Paper # PRMU2005-96
Volume (vol) vol.105
Number (no) 375
Page pp.pp.-
#Pages 6
Date of Issue