R O K E R S
V I S I O N
L A B O R A T O R Y
University of Wisconsin - Madison
Department of Psychology
|Research | Publications | Participate | People|
The stimulus in Figure 3 is compatible with a purely 2D rigid percept. However, after some time, observers predominantly report a 3D percept. For the 3D percept there appears to be a counter-clockwise rotation of the stimulus around its center, in addition to the rotational motion around the center of the image.
We can ask observers to adjust the rotation of the 3D percept around its center, so that no additional apparent rotation remains. This gives us an estimate of the 3D percept generated by the solid line stimulus in Figure 1. The average adjustment is shown below in Figure 4.
In our paper we showed that this adjustment is virtually identical across observers. This is surprising since the stimulus in Figure 1 is compatible with any rotation of the 3D percept around its center. To account for these findings we showed that the estimated percept corresponds closely to a stimulus that minimizes motion slowness and smoothness.
Although it appears there is no additional rotation around the contour in Figure 4, the contour is rotating relative to the stimulus. This can be checked by explicitly tracking points on the contour.
Note that in Figure 4 there is a small discontinuity of the motion due to the looping of the movie. In our experiments this discontinuity was not present.