Spatial and Colour Opponency in Anatomically Constrained Deep Networks

TL;DR

This paper explores how anatomically constrained deep neural networks develop spatial and colour opponency, revealing emergent properties similar to biological visual systems and offering insights into neural processing of colour and spatial information.

Contribution

The study demonstrates the emergence of opponent cells in deep networks constrained by biological principles, linking neural architecture to functional colour and spatial processing.

Findings

Emergence of single and double opponent cells in networks

Distribution of opponent cells varies with retinal bottleneck constraints

Provides new tools for understanding deep learning black box

Abstract

Colour vision has long fascinated scientists, who have sought to understand both the physiology of the mechanics of colour vision and the psychophysics of colour perception. We consider representations of colour in anatomically constrained convolutional deep neural networks. Following ideas from neuroscience, we classify cells in early layers into groups relating to their spectral and spatial functionality. We show the emergence of single and double opponent cells in our networks and characterise how the distribution of these cells changes under the constraint of a retinal bottleneck. Our experiments not only open up a new understanding of how deep networks process spatial and colour information, but also provide new tools to help understand the black box of deep learning. The code for all experiments is avaialable at \url{https://github.com/ecs-vlc/opponency}.