Abstract
The relative input strength of signals from the two eyes has recently been quantified in various ways (e.g., Hess et al., 2010; Ooi et al., 2013; Ding et al., 2013). Ooi et al. (2013) showed that an early stimulus to one eye increased that eye's weight during binocular combination, as revealed by changes in the interocular contrast balance to achieve equal effectiveness in binocular rivalry tasks. However, it is not clear how general this effect is within the hierarchy of visual perception. We asked whether early (300 or 50 ms) onset of contrast in one eye would reduce the contrast needed to detect motion direction in a dichoptic random-dot kinematogram (RDK), to determine the specificity of this enhancement. In a 2AFC motion discrimination task, coherently moving signal dots and randomly moving noise dots were displayed to the two eyes respectively. A 3-down 1-up staircase controlled the interocular signal-to-noise contrast ratio and measured the threshold at which the human observer reported the moving direction of the signal dots at 82% correct. A static monocular "cue" consisted of a group of high-contrast dots that was briefly displayed before the dichoptic RDK. The cue was displayed to the left or the right or neither eye when signal dots were in the left or the right eye; cue color was grey or else red to make the cue more salient. We tested both amblyopes and normally sighted observers. We found no systematic change in the interocular balance as a function of cuing. The cued eye showed no advantage in global motion perception. We suggest that (1) binocular processes in rivalry and RDK tasks do not inherit the same interocular balance, and/or (2) the early-cue effect requires that test stimuli not contain too much motion energy of their own.
Meeting abstract presented at VSS 2014