Rational Value Sequence based Singular Value Decomposition for Facial Recognition

Various prominent feature points of human faces play a critical role for designing a system that can perform facial recognition. Involvement of facial expression as precision parameter exponentially increases the challenges in developing a face recognition system. There exist abundant volume of studies in the area of face recognition, but majority of the techniques are either computationally not viable or uses complex algorithms. Hence, the prime aim of this paper is to introduce a very simple and cost effective solution for face recognition system from facial expressions by using two techniques i.e. Singular Value Decomposition (SVD) and Rational Value Sequence based Singular Value Decomposition (RSVD). Results were evaluated for both the cases to find RSVD approach as a better alternative to SVD approach for face recognition in terms of preciseness and computational effectiveness.