Abstract
We explored the effect of depth perception in biological motion (BM) perception. The stimuli were the point-light walkers comprised of dynamic random dot stereogram (dot density was 50 %). Joints of the walker were either in front (front-walker condition) or behind (behind-walker condition) of the background with the binocular disparity (Disparity-defined Biological motion, DB). In the control condition, joints were defined by the dot density or mean luminance (25 % density, Luminance- defined Biological motion, LB). As another control condition, we examined the performance to detect the movement of a single joint defined by the binocular disparity (Disparity-defined Single motion, DS). Subjects were asked to judge the direction of walker's articular movement (forward or backward walking). In DB, the performance was significantly higher (67 %) than chance level when the walker was in front, but that was chance level when the walker was behind of the background. The performances for two control conditions (LB and DS) were about 100 %. In the next experiment, we investigated the perception of BM when the LB and DB were superimposed while their directions of articular movements were opposite. The luminance contrast of walker was 5, 10, 15, or 20 %. Subjects were asked to judge the direction of LB, but they were unaware that the articular movement of DB was opposite to the LB. They correctly perceived the direction of LB when the luminance contrast was strong. However, when the luminance contrast was weak, they perceived the direction of DB as LB direction in the front-walker condition. In the behind-walker condition, the perception of LB persisted. These results indicate that there is a common process of BM perception independent of whether luminance or disparity defines walker, and it processes only the objects perceived as in front.