Learning Generative Models via Discriminative Approaches

Abstract

Generative model learning is one of the key problems in machine learning and computer vision. Currently the use of generative models is limited due to the difficulty in effec- tively learning them. A new learning framework is proposed in this paper which progressively learns a target genera- tive distribution through discriminative approaches. This framework provides many interesting aspects to the liter- a ture . From the generative model side: (1) A reference distribution is used to assist the learning process, which
removes the need for a sampling processes in the early stages. (2) The classification power of discriminative ap- proaches, e.g. boosting, is directly utilized. (3) The abil- ity to select/explore features from a large candidate pool allows us to make nearly no assumptions about the train- ing data. From the discriminative model side: (1) This framework improves the modeling capability of discrimina- tive models. (2) It can start with source training data only and gradually “invent” negative samples. (3) We show how
sampling schemes can be introduced to discriminative mod- els. (4) The learning procedure helps to tighten the decision boundaries for classification, and therefore, improves ro-bustness. In this paper, we show a variety of applications including texture modeling and classification, non-photo- realistic rendering, learning image statistics/denoising, and face modeling. The framework handles both homogeneous patterns, e.g. textures, and inhomogeneous patterns, e.g. faces, with nearly an identical parameter setting for all the
tasks in the learning stage.