Purchase this article with an account.
Jon K. Grossmann, Allan C. Dobbins; Rotating Kinetic Dot Patterns Stabilize Perceptual Dominance During Binocular Rivalry. Journal of Vision 2004;4(8):247. doi: https://doi.org/10.1167/4.8.247.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
A multistable stimulus can be locked into one interpretation for extended periods by periodically blanking the stimulus. Using this interrupted method of presentation, the rate of perceptual switching can be slowed for both ambiguous figures and binocularly rivalrous stimuli (Leopold, et al., 2002, Nat. Neurosci). We investigated whether stabilization of a binocularly rivalrous stimulus reflects selection of the same eye's input over time, or stabilization of the dominant pattern. Observers were presented dichoptically with either circular patches of dots translating in opposite directions, orthogonal gratings, or opaque random dot covered objects rotating in opposite directions in each eye. The stimuli were repeatedly turned on for 1.0–1.5s and off for 2.0–2.5s while observers continually reported which of the two rival patterns was perceptually dominant. In order to dissociate eye dominance from pattern dominance, the stimuli were sometimes swapped between the eyes between successive stimulus ON times. We found that when viewing orthogonal gratings or opposite dot motion, the input from one eye would stabilize over time, so the observer reported seeing whichever stimulus was presented to that eye. In contrast, when viewing rivalrous rotating objects, most observers continuously reported seeing the same direction of rotation, even when the stimuli were switched between the eyes. Since the neural substrate for rotation in depth is likely to be in higher visual areas and not V1, input selection must not be decided entirely within V1. These results support either a multistage model of rivalry, or the idea that feedback from higher level visual areas can bias input selection in early visual areas such as V1. For example, input selection may be biased in this case by feedback from an object-based representation to favor a particular sense of rotation.
This PDF is available to Subscribers Only