Item

Abstract

Real and illusory contour cues are used under different contexts and can interfere with one another. We have shown that illusory contour percepts are degraded by the presence of real lines (DillenburgerRoe 2004). Optical imaging and single unit data in the primate suggest suppressive feedback interactions from illusory contour cells in V2 to similarly oriented cells in V1 (Ramsden etal 2001). We hypothesize that interference of illusory contour percepts occurs via competition between feedforward (real) vs feedback (illusory) influences. We further test this hypothesis by examining the timing of this interference using psychophysical methods. We expect an early (feedforward) enhancement and a late (feedback) interference of illusory percept by real lines.
We measured the perceptual strength of Kanizsa-type illusory contours in a 2AFC shape discrimination task using real contours as backward masking stimuli. Illusory contours were presented for 50 msec followed by a real line superimposed and parallel to the illusory contour. Masks were presented for 100 msec with varying SOAs at different super- and sub-threshold contrasts. We used a blank and orthogonally oriented real lines as comparison masks.
We found that the influence of real lines on illusory contour perception changes over 0-150 msec post-stimulus onset. Early in illusory contour perception, real lines parallel to the illusory contour enhance the perceptibility of the illusory contour (SOA 50-100 msec). At later times (SOA 125-150 msec), parallel real lines interfere with illusory contour percept.
Our results support an early feedforward-driven enhancement of illusory contours by real lines, perhaps mediated by enhancing illusory contour cell response in V2. Stimulation by real lines (feedforward) at a later time point of illusory contour processing competes with the feedback-driven suppression normally needed for illusory percept consolidation.