Abstract
First-order motion is assumed to be detected by “motion energy” or equivalently “Reichardt Detectors.” The output of such detectors is “motion strength”, a quantity that is proportional to the square of the contrast of a moving sinewave grating. When two sine gratings of possibly different contrasts and orientations move together (a plaid), do the component motion strength vectors sum to determine perceived direction of the plaid? To investigate this question, we did a series of experiments using two component gratings with various spatiotemporal frequencies, different angles between component gratings, different durations of presentation, and a wide range of contrasts for the component gratings. In plaid sinewave stimuli in which the contrasts of the components differed by less than a factor of about 4, the two components contributed equally to the combined motion. When the contrast of one component grating exceeded the other by more than 4, it masked the motion of the lower-contrast grating and only the larger component contributed to the combined motion. The results indicate that motion-strength vectors are not what is summed in plaid motion perception.
Air Force Office of Scientific Research, Life Sciences, Grant FA9550-04-1-0225