Arthur J. Chick Professor of Electrical Engineering and Computer Science
Professor Emeritus of Vision Science and Optometry
Computer vision and computational modeling of human vision
I am interested in constructing computational models of early visual processes such as texture perception, stereopsis and motion processing. This modeling is to be done in the context of available psychophysical and physiological data – it is unlikely that models designed purely on a priori considerations will either be correct or be taken seriously by experimentalists. Computational models force us to make explicit various assumptions; they provide important tests of internal consistency and can also be simulated for robustness.
Our recent work has been on the problems of texture discrimination and stereopsis. We have been able to construct a simple, physiologically plausible model of texture discrimination, which has been able to quantitatively explain psychophysical data on the degree of discriminability of different texture pairs. In stereopsis, we have developed a novel computational model which enables the treatment of position, orientation, and spatial frequency disparity in a single framework.