Color Visual Illusions: A Statistics-based Computational Model

TL;DR

This paper introduces a data-driven computational model that explains and generates color visual illusions by analyzing patch likelihoods in large image datasets, unifying lightness and color illusions.

Contribution

It presents a novel, statistics-based tool and model that explain visual illusions and can generate illusions in natural images, advancing input-driven understanding in vision science.

Findings

The model successfully explains existing visual illusions.

It can generate new illusions in natural images.

The approach unifies lightness and color illusion explanations.

Abstract

Visual illusions may be explained by the likelihood of patches in real-world images, as argued by input-driven paradigms in Neuro-Science. However, neither the data nor the tools existed in the past to extensively support these explanations. The era of big data opens a new opportunity to study input-driven approaches. We introduce a tool that computes the likelihood of patches, given a large dataset to learn from. Given this tool, we present a model that supports the approach and explains lightness and color visual illusions in a unified manner. Furthermore, our model generates visual illusions in natural images, by applying the same tool, reversely.