CAREER: Stochastic Modeling and Computing of Visual Patterns: From Descriptive to Generative Methods

University Of California-Los Angeles, Los Angeles CA

Investigators

Abstract

The project focuses on integration of two learning paradigms for modeling stochastic visual patterns, and on developing a computational paradigm for effective stochastic inference. The work will strengthen the move from descriptive to generative models. Algorithms for learning and inference stronger than those existing today will be developed. A few order of magnitude improvement is expected for the mew methods. There is a good education program that includes improvement in the curriculum and design of new courses. Strong collaboration with industry and governmental labs will provide students with interesting experience.