Face Recognition using FSS-DSOP for Small Sample Size Problem with Illumination Variations

In this paper we propose a method termed as Face-SpecificSubspace DCT Sign Only Product (FSS–DSOP), for facerecognition. The proposed method is based on Face-SpecificSubspace (FSS) technique using Discrete Cosine Transform (DCT)sign only product as similarity measure. The proposed approachaims to improve the recognition rate of the FSS technique undersmall size scenario with illumination variations by making use ofDCT sign-only inner product similarity measure for classification.Results obtained from a set of experiments using small training setssampled from illumination subset of extended Yale-B and PIE facedatabase indicate that the proposed methodology achieves betterrecognition rate when compared to FSS.