Abstract
Crowding refers to a subject's difficulty in identifying a target in the presence of distracters. As a first attempt towards identifying the neural mechanism of crowding, we investigated perceptual crowding using a random-dot kinematogram. Subjects were required to report the direction of moving dots (up or down) within a center patch (3 deg) of a center/surround display presented 10 degrees to the left of fixation, and to ignore the dots in the surround. Motion coherence of the dots in the center patch, as well as surround size (3, 4.5, 6, 9, 12 deg) varied randomly across trials. Motion coherence of the surround was always set at 0 percent. For each of 11 subjects, we calculated direction discrimination thresholds (at 82% correct) at each surround size. Consistent with crowding, thresholds increased when surround size was 4.5 and 6 degrees, compared to those with no surround. Surprisingly, however, thresholds started to decrease when surround size was 9 and 12 degrees, relative to 6 degrees. Our results show that the spatial resolution of motion direction discrimination improves when the area we have to ignore exceeds a defined size. We propose that the improvement of spatial resolution with increased surround noise can be explained by the inhibitory receptive field of a motion area such as the middle temporal area (MT).