hide
Free keywords:
-
Abstract:
For binocular depth perception, right and left eye information must be compared, and the
difference taken. When presented with anticorrelated displays (elements in one eye match those
of opposite contrast polarity in the other eye) there is little or no consistent depth perception.
Here we studied a situation where information from the two eyes would ideally be averaged for
subsequent visual processing. Texture/density perception was explored using a relative numerosity
task. Stimuli were composed of black and white dots, presented for 400 ms. We compared rela-
tive numerosity judgements for two conditions: (i) binocular and (ii) binocularly anticorrelated
stimuli. First, we compared performance for discrimination where both test and comparison stimuli
were either binocular or anticorrelated. We hypothesized that precision using the anticorrelated
stimuli could be compromised by binocular rivalry. The JNDs for the binocular comparison were
significantly lower than for the anticorrelated ones. Second, we compared performance where
standard and test were presented under the two different conditions. We expected that a rivalrous
percept could lead to perceiving up to twice as many dots in the anticorrelated condition. This
did not occur (there was no significant bias), suggesting that observers saw the same number of
dots in both conditions. For the numerosity task, binocularly anticorrelated and correlated dots
appear to be combined between the eyes in a similar fashion. For binocular disparity tasks, depth
perception is not achieved for anticorrelated stimuli. We find here that performance in a relative
numerosity task is not adversely affected by anticorrelation between right and left eye.
