Face recognition using principal component analysis

Abstract

This paper discusses the application of Principal Component Analysis (PCA) in face recognition systems. PCA subspace models are commonly used to perform image dimensionreduction before the input of the classifier. More recently, PCA subspace models are estimated for one face and the comparison of models via a subspace distance allows face identification. Both strategies of applying PCA were compared for a repository of faces of famous people in uncontrolled poses.