Face verification through tracking facial features

We propose an algorithm for face verification through tracking facial featu res by using sequential importance sampling. Specifically, we first formula te tracking as a Bayesian inference problem and propose to use Man kov chai n Monte Carlo techniques for obtaining an empirical solution. A reparameter ization is introduced under parametric motion assumption, which facilitates the empirical estimation and also allows verification to be addressed alon g with tracking. The facial features to be tracked are defined on a grid wi th Gabor attributes (jets). The motion of facial feature points is modeled as a global two-dimensional (2-D) affine transformation (accounting for hea d motion) plus a local deformation (accounting for residual motion that is due to inaccuracies in 2-D affine modeling and other factors such as facial expression). Motion of both types is processed simultaneously by the track er: The global motion is estimated by importance sampling, and the residual motion is handled by incorporating local deformation into the measurement likelihood in computing the weight of a sample. Experiments with a real dat abase of face image sequences are presented. (C) 2001 Optical Society of Am erica.