The general goal of the research is to understand and model the perception of visual patterns by normal, human, adult viewers. The proposed experiments investigate how the visual system combines the information provided by the different components of a multicomponent spatial pattern. The individual components are spatial Gabor wavelets which, existing physiological and psychophysical research indicates, are processed separately at early cortical stages. Patterns are formed by spatially and temporally superimposing components which differ enough in the Fourier domain (e.g. 90 deg difference in orientation, 2 octave difference in spatial frequency) to assure separate initial processing.
The specific aims are to (1) test and elaborate current models of nonlinear interactions, specifically contrast gain control mechanisms; and (2) define the characteristics of second-order combining networks which combine information from different components in order to represent object features such as textures, edges, and motion. The proposed experiments are psychophysical and the task of the viewer is always to discriminate between two very similar patterns. Discrimination under masking is used to isolate and evaluate nonlinear processes. Combining processes are isolated by holding the number of components constant while varying the number of components which provide cues to discrimination and varying the way in which the cues are combined (configuration effect). The results of the proposed studies will be useful in designing and interpreting physiological studies of post-V1 cortical regions, e.g. V4, and may lead to improved understanding of visual deficits, such as dyslexia, in the perception of multi-component patterns.
