Lightness from contrast: a selective integration model

Area of Science:

• Visual perception
• Computational neuroscience
• Psychophysics

Background:

• Perceived lightness traditionally linked to luminance ratios of adjacent regions.
• Retinal mechanisms like adaptation and lateral inhibition process luminance contrast.
• Simple models fail to explain complex lightness illusions, such as White's illusion.

Purpose of the Study:

• To present a novel computational model for simulating perceived lightness.
• To extend existing edge integration and contrast/filling-in models.
• To account for a wider range of visual phenomena beyond simple contrast ratios.

Main Methods:

• Developed a selective integration model (SIM).
• Built upon established edge integration and contrast/filling-in frameworks.
• Tested the model's ability to simulate various visual illusions and effects.

Main Results:

• The SIM successfully simulates White's illusion, where perceived lightness contradicts local contrast ratios.
• The model accounts for phenomena like the Benary Cross, shading, transparency, and Gelb induced contrast.
• An independently derived variant of a depthful White's illusion was also simulated.

Conclusions:

• The selective integration model (SIM) offers a more comprehensive explanation for perceived lightness across diverse visual scenes.
• The model's success in simulating complex illusions suggests a refined understanding of visual processing.
• The SIM framework can inspire the creation of novel visual stimuli and further research.