Presentation 2013-09-02
Gaze Point Prediction with Gap Structure Models
Ryo YONETANI, Hiroaki KAWASHIMA, Takashi MATSUYAMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) When we are watching videos, there exist spatiotemporal gaps between salient regions that attract visual attention and actual points of gaze. These gaps result from reaction delays or anticipation in eye movements toward dynamic changes of the salient regions (saliency dynamics). In this paper, we exploit the underlying structures of the gaps for gaze point prediction from videos. The proposed method involves the relationships between the gaps and the saliency dynamics by introducing localized primitive patterns of salient regions for the description of saliency dynamics. It enables us to learn ``what types of saliency dynamics can be observed around the points of gaze'' in a data-driven fashion for the gaze point prediction.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) saliency dynamics / eye movements / non-negative matrix factorization
Paper # PRMU2013-41,IBISML2013-21
Date of Issue

Conference Information
Committee PRMU
Conference Date 2013/8/26(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) Gaze Point Prediction with Gap Structure Models
Sub Title (in English)
Keyword(1) saliency dynamics
Keyword(2) eye movements
Keyword(3) non-negative matrix factorization
1st Author's Name Ryo YONETANI
1st Author's Affiliation Graduate School of Informatics, Kyoto University()
2nd Author's Name Hiroaki KAWASHIMA
2nd Author's Affiliation Graduate School of Informatics, Kyoto University
3rd Author's Name Takashi MATSUYAMA
3rd Author's Affiliation Graduate School of Informatics, Kyoto University
Date 2013-09-02
Paper # PRMU2013-41,IBISML2013-21
Volume (vol) vol.113
Number (no) 196
Page pp.pp.-
#Pages 6
Date of Issue