Title: “Perceptual and cortical mechanisms of surface lightness representation” by Dr. Michael E. Rudd
Department of Physiology and Biophysics University of Washington School of Medicine
Seattle, WA 98195 USA
Date: Wednesday, April 13
Time: 12:40 p.m.
Location: Room A-130 FEASS
I will describe some perceptual experiments designed to test current theories of lightness perception (the perception of the gray levels of achromatic material surfaces). The results challenge some leading lightness theories, but they are consistent with a computational model that I have proposed regarding the neural formation of surface lightness representations within the ventral stream of visual cortex. According to the model, information about local oriented contrast in the retinal image is first encoded by oriented contrast detectors in early visual cortex, then summed across space at a higher cortical stage. Our lightness judgments correspond to the output of this higher stage. Additional cortical mechanisms of attentional gain control and Gestalt grouping together determine how local contrasts are selected for spatial summation. I will describe how instructions to an experimental observer to interpret a particular luminance edge in the retinal image to be the result of either of a surface color change, or a shadow, changes both the observer’s lightness judgments and the cortical computations in the model observer. The lightness computations embodied by model make sense in terms of the ecological need to discount variations in illumination in the natural world so that the brain can create and maintain stable neural representations of visual objects in the environment. As a sidelight, I will show how the model explains a new lightness illusion—the Phantom Illusion—in which surrounding a dark or light patch with a properly oriented luminance grating can reverse the apparent contrast of the patch from dark to light, and vice versa. Although this illusion runs counter to intuition, as well as to much of the thinking community of lightness researchers, it is explained by the cortical model on the basis of the same principles that serve to support lightness constancy in natural ecological vision.