| Paper: | MLSP-P5.5 | ||
| Session: | Image and Video Processing | ||
| Time: | Thursday, May 20, 13:00 - 15:00 | ||
| Presentation: | Poster | ||
| Topic: | Machine Learning for Signal Processing: Signal detection, Pattern Recognition and Classification | ||
| Title: | FACE RELIGHTING FOR FACE RECOGNITION UNDER GENERIC ILLUMINATION | ||
| Authors: | Laiyun Qing; Chinese Academy of Sciences | ||
| Shiguang Shan; Chinese Academy of Sciences | |||
| Xilin Chen; Chinese Academy of Sciences / Harbin Institute of Technology | |||
| Abstract: | The performance of current face recognition systems suffers heavily from the variations of lighting. To deal with this problem, this paper presents a novel face normalization approach by relighting faces under normal illumination based on harmonic images. Benefit from the observations that the human faces share similar 3D shape and the albedo of the face surface is quasi-constant, given a face image, we first estimate the spherical harmonics coefficients of the illumination environment under which the images is taken. Then, the face image is normalized by relighting it to canonical illumination based on the illumination ratio image. For face recognition purpose, two kinds of canonical illuminations, uniform and frontal point lighting, are considered, among which the former encodes merely texture information, while the latter encodes both texture and shape information. Our experimental results show that the proposed relighting normalization can significantly improve the performance of a face recognition system. | ||
| Back | |||
Home -||-
Organizing Committee -||-
Technical Committee -||-
Technical Program -||-
Plenaries
Paper Submission -||-
Special Sessions -||-
ITT -||-
Paper Review -||-
Exhibits -||-
Tutorials
Information -||-
Registration -||-
Travel Insurance -||-
Housing -||-
Workshops