This Title All WIREs
How to cite this WIREs title:
WIREs Cogn Sci
Impact Factor: 3.476

Binocular rivalry: competition and inhibition in visual perception

Full article on Wiley Online Library:   HTML PDF

Can't access this content? Tell your librarian.

Abstract When the brain is presented with ambiguous visual stimuli supporting two interpretations, perception becomes bistable and alternates over time between one interpretation and the other. This process contains elements of competition (between the rivaling percepts) as well as inhibition, as the percepts are mutually exclusive so that one is always suppressed. This review covers the most widely studied form of bistable perception—binocular rivalry. Suppression in rivalry is covered in detail, including discussion of its general and specific components, its local nature and spatial organization, techniques for quantifying it, and the role of global feedback. The competitive dynamics of rivalry are discussed within the context of the classical ‘adapting reciprocal inhibition’ model of rivalry and recent evidence supporting this model is discussed. This model is contrasted with alternative models based on late competition and with hybrid models. Finally, the role of attention in rivalry is examined and commonalities with other forms of bistable perception are noted. WIREs Cogn Sci 2012, 3:87–103. doi: 10.1002/wcs.151 This article is categorized under: Psychology > Attention

Examples of perceptually ambiguous stimuli. Inspecting any of these figures will elicit perceptual alternations between two roughly equally probable interpretations. For the first two stimuli, there is an ambiguity of perspective that arises from the two‐dimensional rendering of a three‐dimensional form. Over time, the two perspectives or ‘view points’ alternate. The third example shows an instance where the ambiguity concerns the segregation of figure and ground. It is perceived as a vase when the white region is interpreted as figure, or as two faces in profile when the black region is interpreted as figure.

[ Normal View | Magnified View ]

(a) The classical model of rivalry is based on reciprocal inhibition between monocular channels. This model explains how an image becomes suppressed, and the ongoing alternation dynamics are attributed to adaptation occurring within this framework and shifting the balance of inhibition. (b) The method used in a recent study53 to demonstrate a key prediction from adapting reciprocal inhibition: that suppression strength should weaken during a rivalry phase. (c) Data from the method illustrated in (b) showing evidence for adapting reciprocal inhibition. Sensitivity to a contrast increment probe declines over a period of dominance, and correspondingly rises during suppression, and converge just prior to a change of dominance.

[ Normal View | Magnified View ]

Influences of global context on rivalry. (a) When a single element (here in the lower‐right corner) of a global motion stimulus engages in rivalry, it tends to resist suppression and dominate longer than when presented without the global context. This tendency to preserve the global stimulus is attributed to feedback from extrastriate global processing areas modulating the local rivalry process between the two rivalrous gratings. (b) When point‐light walkers are presented in animation, they produce a percept of global form and motion. Point‐light walkers moving in opposite directions will trigger rivalry and also reveal a role for global feedback in local rivalry (see text). (c) Placing a surround around one of the rival stimuli will also modulate rivalry dynamics. A surround with the same orientation will reduce that stimulus' rivalry strength and therefore reduce its predominance, an effect consistent with isosurround suppression effects observed neurophysiologically in primary visual cortex.

[ Normal View | Magnified View ]

Rivalry involves local zones of suppression. (a) When the extent of a rivalry zone is measured as a function of eccentricity, it is found to increase at a similar rate to the cortical magnification function, suggesting it is determined by receptive field size in early cortex. (b) Local rivalry zones interact over space and the alternation dynamics of adjacent rivalry process will tend to correlate. The degree of correlation depends on relative orientation, and is stronger for collinear contours than for oblique and orthogonal contours—a pattern that fits with what is known about lateral connections in primary visual cortex. (c) With large rivaling images, a dominance change will often begin in a single rivalry zone and spread through adjacent regions in a wave‐like manner. These traveling waves of dominance are faster and longer in range along collinear contours than orthogonal contours.

[ Normal View | Magnified View ]

(a) To measure rivalry suppression strength, contrast thresholds are measured for a brief and temporally smoothed probe presented to the suppressed eye. This is compared against the threshold for the dominant eye. Subjects use a key press to indicate when they enter a period of dominance of subject. Psychometric functions may be steeper for the dominant eye, and the dominant threshold will be lower. Ideally, the probe would be a contrast increment of the rival image being measured, and this can be done in the upper or lower half of the image to allow a two‐alternative forced‐choice task. (b) Strength of suppression will vary with the type of rival stimuli. Using the suppression index shown, rivalry suppression is weakest for simple grating stimuli (∼0.3) and stronger for complex objects (∼0.7). Rivalry between complex and simple stimuli also yields weak suppression. A study by Alais and Parker24 showed that rivaling global forms produced no suppression for global motion probes, and vice versa (see text). (c) It is well known that rivalry suppression involves a general and nonselective component (here about 0.3), but recent work suggests that there is also an equivalently sized specific component tuned to the suppressed feature.23

[ Normal View | Magnified View ]

(a) Binocular rivalry is a very actively researched form a bistable perception. It requires presenting separate images to the eyes, usually by means of a mirror stereoscope. Any significant interocular difference in orientation, color, texture, movement, and so on will suffice to trigger rivalry. (b) Binocular rivalry is experienced as a series of irregular perceptual alternations over time as first one image is perceived and then the other. While one image is perceived, the other is suppressed from visual awareness. A given image therefore undergoes periods of dominance and suppression. Two dependent variables that capture rivalry dynamics are predominance (proportion of total viewing time that a given image is dominant), or mean dominance period. (c) All forms of bistable perception produce a skewed Γ‐like distribution when the durations of many dominance periods are pooled. For binocular rivalry, the peak of this distribution typically would be around 1–2 s.

[ Normal View | Magnified View ]

Browse by Topic

Psychology > Attention

Access to this WIREs title is by subscription only.

Recommend to Your
Librarian Now!

The latest WIREs articles in your inbox

Sign Up for Article Alerts