Cortical and Sensory Systems
Brad Motter, PhD,
Research Associate Professor
Cortical Representations in Visual Search
The emphasis in these studies has been to relate visual search performance in primates and humans to the cortical representation of objects in visual space. We have established that search performance—roughly expressed as the ability to “see and find” a target object in a variably cluttered scene—is constrained by distances between objects as they are represented in cortical space rather than the corresponding relationships in external visual space. In addition, a probabilistic model of active visual search has been developed which permits a separation of overall performance into those elements that represent chance and visual attention.
Recent effort has focused on developing an “out-folded” representation of the primary visual cortex (V1) that preserves intrinsic cortical curvature and intracortical distances. The model has been used to calculate distances between objects and their size as represented in cortical space as well as demonstrate that “warped” and variably sized receptive fields in higher cortical visual areas (V4) assume uniform dimension and radial symmetry if “back-projected” onto a curvature preserving model of V1.
The lab continues to study the effects of crowding and scene complexity on visual performance in behaving primates and human subjects. In addition, Dr. Motter maintains an active and supported program of single unit recording studies in cortical visual areas (V4) of primates trained on several visual search tasks. Work continues on developing a standard, computer-based method of “out-folding” primary and secondary visual cortical areas in the primate. The goal is to understand how human observers navigate and analyze complex scenes as well as characterize how normal primate and human visual function relates to brain anatomy and physiology.
—Brad C. Motter.
