Cascaded face recognition system via sparse representation

Facial recognition is considered by many to be one of the key areas in the field of biometrics. It has a few very significant advantages over other modalities including universal acceptability and covert acqusition. On the flip side, it provides some very unique challenges in terms of pose and expression variation etc. These challenges make it difficult to develop a real time facial recognition system that provides a very high accuracy rate. Nontheless a few attempts have been made by researchers to provide a robust real time facial recognition system. Probably, the best and most recent attempt is to represent the whole system via sparse representation. In this paper we extend this work by utilizing the cascaded classifier based approach. The results show that the proposed approach provides improved accuracy while still performing in real time.