Presentation 2000/1/21
Nearest Neighbor Face Recognition with Compensating Lighting Conditions
Nobumasa Yamamoto, Takeshi Shakunaga,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) A nearest neighbor approach is discussed in this paper to realize a robust face recognition free with the lighting conditions. The face eigenspace is constructed from a set of face images which includes 50 persons, each in 24 different lighting conditions. In the eigenspace, we discuss a static framework of the nearest neighbor discrimination, in which three kinds of L2 norms are compared. Then a compensational algorithm is proposed, which consists of an estimation of the lighting condition and an image mapping between lighting conditions. Experimental results show that the compensation with lighting estimation improves the discrimination rate. The proposed algorithm can be applied to both the cases of known and unknown lighting conditions.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) nearest neighbor discrimination / eigenspace / illumination estimation / illumination compensation
Paper # PRMU99-207
Date of Issue

Conference Information
Committee PRMU
Conference Date 2000/1/21(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) Nearest Neighbor Face Recognition with Compensating Lighting Conditions
Sub Title (in English)
Keyword(1) nearest neighbor discrimination
Keyword(2) eigenspace
Keyword(3) illumination estimation
Keyword(4) illumination compensation
1st Author's Name Nobumasa Yamamoto
1st Author's Affiliation Department of Information Technology, Okayama University()
2nd Author's Name Takeshi Shakunaga
2nd Author's Affiliation Department of Information Technology, Okayama University
Date 2000/1/21
Paper # PRMU99-207
Volume (vol) vol.99
Number (no) 575
Page pp.pp.-
#Pages 6
Date of Issue