Image variability due to changes in pose and illumination can seriously impair object recognition. This paper presents appearance-based methods which, unlike previous appearance-b...
Athinodoros S. Georghiades, Peter N. Belhumeur, Da...
In this paper, we propose an ICA(Indepdendent Component Analysis) based face recognition algorithm, which is robust to illumination and pose variation. Generally, it is well known...
Tae-Kyun Kim, Hyunwoo Kim, Wonjun Hwang, Seok-Cheo...
In this paper, we propose two novel methods for face recognition under arbitrary unknown lighting by using spherical harmonics illumination representation, which require only one t...
Evaluations of the state-of-the-art of both academic face recognition algorithms and commercial systems have shown that recognition performance of most current technologies degrad...
In building a face recognition system for real-life scenarios, one usually faces the problem that is the selection of a feature-space and preprocessing methods such as alignment u...