CNN-based models for color vision and visual illusions

Abstract

Cellular neural/nonlinear network (CNN) models are introduced to analyze and explain human early vision color perception and some of the visual illusions. A CNN model of Horn's algorithm is shown and a new lightness-determination method is introduced. CNN models for the brightness, the Mach-band, the Zollner, the cafe-wall, and the Poggendorff illusions are also introduced.