Resampling for face recognition

A number of applications require robust human face recognition under varying environmental lighting conditions and different facial expressions, which considerably vary the appearance of human face. However, in many face recognition applications, only a small number of training samples for each subject are available; these samples are not able to capture all the facial appearance variations. We utilize the resampling techniques to generate several subsets of samples from the original training dataset. A classic appearance-based recognizer, LDA-based classifier, is applied to each of the generated subsets to construct a LDA representation for face recognition. The classification results from each subset are integrated by two strategies: majority voting and the sum rule. Experiments conducted on a face database containing 206 subjects (2,060 face images) show that the proposed approaches improve the recognition accuracy of the classical LDA-based face classifier by about 7 percentages.