Presentation 2021-12-17
Analysis of Recognition Grounds of Models for Social Relationship Recognition Using Spatial-temporal TCAV
Runa Sugiyama, Ying JI, Yu Wang, Jien Kato,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Social relationships between people in videos are difficult to recognize with high accuracy using traditional methods. In order to radically enhance recognition accuracy, it is important to investigate which concepts or attributes are important and thus should be captured by the models. This is especially necessary for deep learning models because of a lack of transparency and interpretability. In this paper, for the task of social relationship recognition, we show how to analyze recognition grounds of video recognition models using concepts, based on spatio-temporal TCAV method.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) TCAV / Decision Grounds Analysis / Relationship Recognition / Video Recognition
Paper # PRMU2021-45
Date of Issue 2021-12-09 (PRMU)

Conference Information
Committee PRMU
Conference Date 2021/12/16(2days)
Place (in Japanese) (See Japanese page)
Place (in English) Online
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair Seiichi Uchida(Kyushu Univ.)
Vice Chair Masakazu Iwamura(Osaka Pref. Univ.) / Mitsuru Anpai(Denso IT Lab.)
Secretary Masakazu Iwamura(NTT) / Mitsuru Anpai(Tottori Univ.)
Assistant Kouta Yamaguchi(CyberAgent) / Yusuke Matsui(Univ. of Tokyo)

Paper Information
Registration To Technical Committee on Pattern Recognition and Media Understanding
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Analysis of Recognition Grounds of Models for Social Relationship Recognition Using Spatial-temporal TCAV
Sub Title (in English)
Keyword(1) TCAV
Keyword(2) Decision Grounds Analysis
Keyword(3) Relationship Recognition
Keyword(4) Video Recognition
1st Author's Name Runa Sugiyama
1st Author's Affiliation Ritsumeikan University(Ritsumeikan Univ.)
2nd Author's Name Ying JI
2nd Author's Affiliation Nagoya University(Nagoya Univ.)
3rd Author's Name Yu Wang
3rd Author's Affiliation Ritsumeikan University(Ritsumeikan Univ.)
4th Author's Name Jien Kato
4th Author's Affiliation Ritsumeikan University(Ritsumeikan Univ.)
Date 2021-12-17
Paper # PRMU2021-45
Volume (vol) vol.121
Number (no) PRMU-304
Page pp.pp.113-117(PRMU),
#Pages 5
Date of Issue 2021-12-09 (PRMU)