Composite binocular perception from dichoptic stimulus arrays with similar ensemble information

Area of Science:

• Neuroscience
• Computational Vision
• Psychophysics

Background:

• Human vision relies on binocular disparity to perceive depth, requiring the brain to match corresponding points in the two retinal images.
• The binocular matching problem is inherently ambiguous, with multiple potential solutions, necessitating strategies for disambiguation.

Purpose of the Study:

• To investigate whether the visual system utilizes regularities among neighboring features (feature ensembles) to aid binocular matching.
• To determine if anomalous feature ensembles lead to unsuccessful binocular matches.

Main Methods:

• Dichoptic stimulus arrays of colored circles were presented to participants.
• The color similarity of stimulus items was manipulated to create probable or unusual ensembles.
• Binocular rivalry was used as a proxy to measure the stability of binocular perception.

Main Results:

• Binocular perception was more stable when stimulus items shared similar colors, forming probable ensembles.
• Unusual ensembles, created by dissimilar colors, induced binocular rivalry, suggesting failed matches.
• The findings indicate that the visual system uses ensemble properties to detect unsuccessful binocular matches.

Conclusions:

• Feature ensembles serve as a strategy for the visual system to resolve ambiguities in binocular matching.
• This ensemble-based strategy helps in identifying correct matches, especially in visually complex scenes with redundant features.
• The study reveals a supplemental mechanism for robust 3D visual perception.