Presentation 2016-01-22
Fine-Grained Classification of Pedestrian using Statistical Information
Yoshihito Kokubo, Yu Wang, Jien Kato, Kenji Mase,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) The purpose of this study is fine-grained classification of sex(2 classes), age(5 classes), weight(3 classes), clothing(4 classes) subcategories. The approach , we represent the relation between the subcategories in the graphical model, assuming that there is a correlation for each subcategories. Furthermore, we incorporated the super resolution and division of the bounding box as a pre-processing. The experiment results indicate that the proposed approach has a better classification accuracy comparing to the state-of-the-art. As a result, we confirmed effectiveness of the proposed approach.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) fine-grained classification / deep learning / super resolution / graphical model
Paper # PRMU2015-124,MVE2015-46
Date of Issue 2016-01-14 (PRMU, MVE)

Conference Information
Committee PRMU / IPSJ-CVIM / MVE
Conference Date 2016/1/21(2days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair Eisaku Maeda(NTT) / / Minoru Kobayashi(Meiji Univ.)
Vice Chair Shuji Senda(NEC) / Seiichi Uchida(Kyushu Univ.) / / Yoshinari Kameda(Univ. of Tsukuba)
Secretary Shuji Senda(Mie Univ.) / Seiichi Uchida(DENSO IT Lab.) / / Yoshinari Kameda(Keio Univ.)
Assistant Kazuaki Kondo(Kyoto Univ.) / Akisato Kimura(NTT) / / Reiko Ueoka(Kyushu Univ.) / Masaaki Iiyama(Kyoto Univ.) / Atsunobu Kimura(NTT)

Paper Information
Registration To Technical Committee on Pattern Recognition and Media Understanding / Special Interest Group on Computer Vision and Image Media / Technical Committee on Multimedia and Virtual Environment
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Fine-Grained Classification of Pedestrian using Statistical Information
Sub Title (in English)
Keyword(1) fine-grained classification
Keyword(2) deep learning
Keyword(3) super resolution
Keyword(4) graphical model
1st Author's Name Yoshihito Kokubo
1st Author's Affiliation Nagoya University(NU)
2nd Author's Name Yu Wang
2nd Author's Affiliation Nagoya University(NU)
3rd Author's Name Jien Kato
3rd Author's Affiliation Nagoya University(NU)
4th Author's Name Kenji Mase
4th Author's Affiliation Nagoya University(NU)
Date 2016-01-22
Paper # PRMU2015-124,MVE2015-46
Volume (vol) vol.115
Number (no) PRMU-414,MVE-415
Page pp.pp.169-173(PRMU), pp.169-173(MVE),
#Pages 5
Date of Issue 2016-01-14 (PRMU, MVE)