Illumination invariant face recognition using linear combination of face exemplars

Facial appearance changes induced by lighting variation cause serious performance degradation in face recognition. Current face recognition systems encounter the difficulty to recognize faces under arbitrary illuminations. In this paper, we propose a new face recognition method under arbitrary lighting conditions, given only a single registered image and training data under unknown illuminations. Our proposed method is based on the exemplars which are synthesized from photometric stereo images of training data and the linear combination of those exemplars are used to represent the new face. We make experiments for verifying our approach and compare it with two traditional approaches. As a result, higher recognition rates are reported in these experiments using the illumination subset of Max-Planck Institute Face Database.