Generative Models for Computer Vision

128 p.Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2002.In order to build robust computer vision algorithms, scene models are necessary that are capable of capturing various aspects of the data at the same time. These models should be fairly simple, but capable of adapting to the data. Flexible models, as defined in the machine learning community, are minimally structured probability models with a large number of parameters that can adapt so as to explain the input data. We describe one possible framework for designing and using flexible models for vision. The framework uses structured probability models to describe causes of variability in the data, exact or variational methods for inference, and an expectation-maximization algorithm for parameter estimation. We show that within this framework, we can perform various vision tasks jointly, such as tracking, recognition, occlusion detection, object stabilization, object removal, and filtering. In fact, in this dissertation we argue that dealing with these tasks jointly is easier than combining individually optimized modules in a typical engineering approach to signal processing.U of I OnlyRestricted to the U of I community idenfinitely during batch ingest of legacy ETD