Pose recognition using mixture of exponential family

Pose recognition has recently become a very hot research topic in computer vision and multimedia information processing. In this paper, we propose a generative model for pose recognition based on mixtures of the exponential family of distributions. The distributions which are considered in this paper are the Multivariate Gaussian (MG), Rayleigh (R), Poisson (P), Bernoulli (B) and Centered Laplacian (CL). The model leads to a generic algorithm that is able to perform learning in a more efficient and flexible manner. Subsequently, it can be used to evaluate the performance of each distribution in the task of pose recognition. The approach is also compared to a discriminative approach (i.e. Support Vector Machine) using the frontal face dataset.