Unsupervised learning on spontaneous retinal activity leads to efficient neural representation geometry

TL;DR

This study models early retinal activity using unsupervised learning on retinal wave movies to enhance neural network representations, revealing that such pre-training improves invariance and feature separation without using natural images.

Contribution

It introduces a novel pre-training approach using simulated and real retinal wave data, demonstrating improved neural network invariance and feature separation capabilities.

Findings

Pre-training on retinal waves enhances object invariance in neural networks.

Networks pre-trained on retinal waves better separate image manifolds.

Retinal wave pre-training improves higher-order feature extraction.

Abstract

Prior to the onset of vision, neurons in the developing mammalian retina spontaneously fire in correlated activity patterns known as retinal waves. Experimental evidence suggests that retinal waves strongly influence the emergence of sensory representations before visual experience. We aim to model this early stage of functional development by using movies of neurally active developing retinas as pre-training data for neural networks. Specifically, we pre-train a ResNet-18 with an unsupervised contrastive learning objective (SimCLR) on both simulated and experimentally-obtained movies of retinal waves, then evaluate its performance on image classification tasks. We find that pre-training on retinal waves significantly improves performance on tasks that test object invariance to spatial translation, while slightly improving performance on more complex tasks like image classification. Notably, these performance boosts are realized on held-out natural images even though the pre-training procedure does not include any natural image data. We then propose a geometrical explanation for the increase in network performance, namely that the spatiotemporal characteristics of retinal waves facilitate the formation of separable feature representations. In particular, we demonstrate that networks pre-trained on retinal waves are more effective at separating image manifolds than randomly initialized networks, especially for manifolds defined by sets of spatial translations. These findings indicate that the broad spatiotemporal properties of retinal waves prepare networks for higher order feature extraction.