ADAPTED GENERATIVE MODELS FOR FACE VERIFICATION

Abstract. It has been shown previously that systems based on local features and relatively complexgenerative models, namely 1D Hidden Markov Models (HMMs) and pseudo-2D HMMs, are suitable forface recognition (here we mean both identification and verification). Recently a simpler generative model,namely the Gaussian Mixture Model (GMM), was also shown to perform well. In this paper we first proposeto increase the performance of the GMM approach (without sacrificing its simplicity) through the use oflocal features with embedded positional information; we show that the performance obtained is comparableto 1D HMMs. Secondly, we evaluate different training techniques for both GMM and HMM based systems.We show that the traditionally used Maximum Likelihood (ML) training approach has problems estimatingrobust model parameters when there is only a few training images available; we propose to tackle thisproblem through the use of Maximum a Posteriori (MAP) training, where the lack of data problem canbe effectively circumvented; we show that models estimated with MAP are significantly more robust and areable to generalize to adverse conditions present in the BANCA database.