Stereopsis is adaptive for the natural environment

Area of Science:

• Neuroscience
• Computational Vision
• Visual Perception

Background:

• Forward-facing eyes in humans and animals provide overlapping visual fields.
• Binocular disparities from these fields enable precise depth estimation.
• Matching corresponding points between retinal images (correspondence problem) is computationally complex.

Purpose of the Study:

• To investigate if the range of disparities in natural environments matches the visual system's search range.
• To determine if natural disparity distributions explain stereogram perception.
• To assess if neural disparity preferences align with natural distributions.

Main Methods:

• Simultaneous measurement of binocular eye position and 3D scene geometry during natural behaviors.
• Analysis of the distribution of naturally occurring binocular disparities.
• Comparison of natural disparity distributions with the visual system's search range and neural preferences.

Main Results:

• The distribution of disparities encountered in natural environments closely matches the visual system's limited search range.
• This natural distribution explains the perception of certain ambiguous stereograms.
• Disparity preferences recorded from macaque cortical neurons are consistent with these natural distributions.

Conclusions:

• The visual system's disparity search range is adapted to the statistics of the natural environment.
• This adaptation is crucial for efficient and accurate depth perception.
• Findings support a model where neural processing is tuned to ecologically relevant visual input.