Abstract
A Glass pattern consists of randomly distributed dot pairs, or dipoles, whose orientation is determined by a geometric transform, which defines the global percept that would be perceived by an observer. The perception of Glass patterns involves a local process to associate dot pairs into dipoles and a global process to group the dipoles into a global structure. To further clarify the relationship between the local and global processes, we used patterns consisted of tripoles, or groups of three dots, as stimuli. In each tripole, three dots formed the vertex of an equilateral triangle with the anchoring dot pointing toward the center of the display. Grouping the anchoring dot with one of the context dot (D1) would lead to the percept of a clockwise spiral while the other dot (D2), a counterclockwise spiral. We manipulated the contrasts of the two context dots and measured the percentage of a participant judging the patterns as clockwise (CW) or counter-clockwise (CCW) at various combinations of contrast of the dots. When the contrast of D1 increased, the probability of perceiving CW first increased then decreased. Such inverted-V shape function cannot be accounted for by the energy model for form perception (Prazdny, 1984), which would predict a monotonic increasing function. The peak D1 contrast increased as D2 contrast, inconsistent with the similarity theory, which predicts that the peak would always occur when the D1 and the anchor have the same contrast. Our result was well fit by a contrast normalization model, in which the response to local dipoles is the sum of dot contrast raised by a power and divided by an inhibition signals driven by all dots in the tripole. The local responses are then pooled to determine global percept. Our model potentially reconciles the long disputed contradiction between the two previous models.
Meeting abstract presented at VSS 2016