Face recognition: a review and comparison of HMM, PCA, ICA and neural networks

This paper describes the review-based comparison and implementation of face recognition. The implementation results described use principal component analysis (PCA) as a preprocessing step. Different authors have used PCA for reducing the dimensions of data. In this paper we use PCA as a preprocessor of ICA (independent component analysis). The accuracy and efficiency of the system depend upon the situation where the system is to be used.