The empirical characteristics of human pattern vision defy theoretically-driven expectations

Area of Science:

• Visual perception
• Computational neuroscience
• Image processing

Background:

• Contrast is fundamental to image perception.
• Existing models predict broader visual tuning at low contrast.
• Biological vision models need to incorporate contrast's impact.

Purpose of the Study:

• Investigate human visual system's response to varying contrast levels.
• Challenge current theoretical and computational models of vision.
• Characterize visual processing at low contrast thresholds.

Main Methods:

• Empirical measurements of human contrast discrimination.
• Analysis of visual system's linearity and variability.
• Investigating neural circuit architectures.

Main Results:

• Human vision sharpens, not broadens, as contrast nears threshold.
• Increased internal variability observed at low contrast.
• Transition from linear (high contrast) to nonlinear (low contrast) processing.

Conclusions:

• Empirical findings challenge optimality-based interpretations of human vision.
• Intrinsic architectural features, not optimality, constrain visual processing.
• Current computational models inadequately explain basic visual operations.