Feb. 17, 11:45, Wei Ji Ma. Optimality and Probabilistic Computation in Visual Categorization

Perception and neurons

Qasim Zaidi

Graduate Center for Vision Research

State University of New York

We present three examples where in vivo electrophysiological recordings from neurons have helped us to resolve long-standing conundrums in human visual perception: 1.    We show that the greater salience of darks versus lights, the irradiation illusion, the perceptual distortion of sine-wave gratings, and the higher spatial resolution for darks, can all be explained by a neuronal nonlinearity in the early visual pathway: neurons driven by darks (OFF channel) increase their responses roughly linearly with luminance decrements, whereas neurons driven by lights (ON channel) saturate their responses with small increases in luminance 2.    By measuring responses of retinal ganglion cells to stationary luminance and chromatic edges, and to edges undergoing transient displacements simulating fixational eye-movements, we show that psychophysical effects of context on temporal sensitivity can be explained by transient neuronal responses to displaced edges, but not by lateral inhibitory effects on responses to stationary edges. 3.    We identify the neural locus of color afterimages by using a time-varying method for evoking after-images to show that all three classes of retinal ganglion cells exhibit subtractive adaptation to prolonged stimuli, with much slower time-constants than those expected of photoreceptors.  At the cessation of the stimulus, ganglion cells generate rebound responses that provide afterimage signals for later neurons. These neural processes also serve more general purposes.  Natural images have more darks than lights and the neuronal machinery in the visual pathway allocates a greater linear range to measure variations in darkness than lightness. Through eye-movements, the transformation of sharp spatial edges to transient retinal responses provides the neuronal substrate for detecting chromatic and luminance edges in static natural scenes.  The subtractive retinal adaptation counters the effects of slow changes in ambient illumination to maintain perceptual stability.