Specific and general contour adaptation effects found using a selective adaptation paradigm.

Abstract

There is evidence for discrete property analyzers in mammalian visual systems. Research has indicated that prolonged stimulation of such units reduces their sensitivity to subsequent stimulation. Psychophysical studies have employed this effect, terned selective adaptation, to study feature extractors in the human visual system. The purpose of this study was to determine the role of density and deviation in the adaptation figure on the recognition thresholds of simple and complex text figures. A selective adaptation paradigm was employed. A strict property analyzer model suggests that increases in density, deviation, or complexity should lead to an increased recognition time for the test figures. This was not confirmed. The complexity of the test figure had no effect on its recognition time. Both increased density and deviation did have an effect on the recognition times of the test figures. The results thus suggest that contour adaptation involves at least two processes: a general, probably peripheral effect due to the fatigue of visual receptor units and a more specific effect generated by the similarity between test figure and adaptation contours, independent of the site of stimulation.