Geodesic Illumination Basis: Compensating for Illumination Variations in any Pose for Face Recognition

This paper proposes a new model of illumination variations of the object appearance, culled the geodesic illumination basis model. It calculates poseindependent illumination buses on a 30 model, and these buses are warped into view-dependent buses in any pose. We experimentally evaluate how many illumination samples and buses are necessary, and show that our model can compensate for any illumination vutiutions in any pose. A face recognition system incorporating our proposed model is constructed, and its performance is tested using a database of 30 models and test images of-42 individuals captured in drastically diffeting pose and illumination conditions. Our system achieves a first-choice success ratio of 97.3% when the position and pose of the target face are known.