Foveated Retinotopy Improves Classification and Localization in Convolutional Neural Networks

TL;DR

Embedding a biologically inspired foveated retinotopic transformation in CNNs enhances their robustness to scale and rotation, improves object localization, and integrates prior geometric knowledge inspired by biological vision.

Contribution

This work introduces a foveated retinotopic preprocessing layer in CNNs, demonstrating improved robustness and localization by incorporating biological visual principles.

Findings

Retaining accuracy while improving robustness to scale and rotation.

Using fixation-point shifts as a proxy for saliency maps.

Enhancing object localization through retinotopic encoding.

Abstract

From falcons spotting preys to humans recognizing faces, rapid visual abilities depend on a foveated retinal organization which delivers high-acuity central vision while preserving low-resolution periphery. This organization is conserved along early visual pathways but remains underexplored in machine learning. Here we examine how embedding a foveated retinotopic transformation as a preprocessing layer impacts convolutional neural networks (CNNs) for image classification. By applying a log-polar mapping to off-the-shelf models and retraining them, we retain comparable accuracy while improving robustness to scale and rotation. We show that this architecture becomes highly sensitive to fixation-point shifts, and that this sensitivity yields a proxy for defining saliency maps that effectively facilitates object localization. Our results show that foveated retinotopy encodes prior geometric knowledge, offering a solution to visual-search and enhancing both classification and localization. These findings connect biological vision principles with artificial networks, pointing to new, robust and efficient directions for computer-vision systems.