Motion induction or assimilation in depth in a frontoparallel moving display? The importance of depth cues

Abstract

Induced motion in the frontoparallel plane occurs when viewing displays composed of a vertically oscillating target dot enclosed within a horizontally oscillating inducer frame. In such displays, the dot appears to be moving diagonally, as would be predicted from subtracting the frame's motion from that of the dot. Separating the frame and the dot in depth has only a limited influence on induced motion, suggesting that the 3D layout of the visual scene does not affect frontoparallel induced motion. On the other hand, previous experiments have demonstrated the possibility of induced motion in depth when the dot and frame move along that dimension. Here, we demonstrate in two experiments that a component of motion in depth emerges in a frontoparallel moving display but where a degree of slant is imparted to the inducer frame, using either monocularly or binocularly viewed perspective, or disparity as depth cue. Our results show that the target dot is perceived as moving in depth either along the surface of the frame (3D motion assimilation) or against it (3D induced motion) when the dominant depth cue is perspective or disparity, respectively. We propose a model of motion perception which predicts this pattern of results based on the hypothesis that relative depth is computed only locally, consistent with an earlier model formulated for the case of static displays.

Supported in part by CELEST, an NSF Science of Learning Center (SBE-0354378 and OMA-0835976) and by the SyNAPSE program of DARPA (HR0011-09-03-0001).