Neurocomputational Lightness Model Explains the Appearance of Real Surfaces Viewed Under Gelb Illumination

Area of Science:

• Visual Perception
• Computational Neuroscience
• Color Science

Background:

• Visual perception's role in evaluating material surface properties, including color and lightness.
• Importance of surface color in conveying ecological and social information.
• Existing models of lightness perception and their limitations.

Purpose of the Study:

• To further develop and apply a neural model for human lightness perception.
• To simulate and quantitatively analyze lightness judgments of achromatic surfaces under spotlight illumination.
• To explain lightness constancy failures and dynamic range compression phenomena.

Main Methods:

• Utilized a neural model based on ON- and OFF-cell responses to luminance ratios.
• Characterized cell response properties using physiologically motivated equations.
• Employed an edge integration process, similar to the retinex model, for lightness computation.

Main Results:

• The model quantitatively accounts for lightness judgments under spotlight illumination.
• ON-cells showed compressive power law response; OFF-cells showed linear response.
• Model successfully predicted lightness constancy failures, dynamic range compression, and compression releases.

Conclusions:

• The developed neural model provides a robust framework for understanding human lightness perception.
• The model's parameters, including polarity- and distance-dependent factors, explain complex perceptual phenomena.
• This work advances computational models of visual processing and color science.