Neural information coding for efficient spike-based image denoising

TL;DR

This paper explores the use of Spiking Neural Networks with Leaky Integrate and Fire neurons for image denoising, aiming to match the performance of traditional deep CNNs while significantly reducing computational requirements.

Contribution

It introduces a formal analysis of neural coding schemes in SNNs and demonstrates their potential for efficient image denoising compared to classical CNNs.

Findings

SNNs with LIF neurons achieve competitive denoising performance.

SNNs reduce computational load compared to traditional CNNs.

Rate-coding mechanisms are compared with LIF neurons for information processing.

Abstract

In recent years, Deep Convolutional Neural Networks (DCNNs) have outreached the performance of classical algorithms for image restoration tasks. However most of these methods are not suited for computational efficiency and are therefore too expensive to be executed on embedded and mobile devices. In this work we investigate Spiking Neural Networks (SNNs) for Gaussian denoising, with the goal of approaching the performance of conventional DCNN while reducing the computational load. We propose a formal analysis of the information conversion processing carried out by the Leaky Integrate and Fire (LIF) neurons and we compare its performance with the classical rate-coding mechanism. The neural coding schemes are then evaluated through experiments in terms of denoising performance and computation efficiency for a state-of-the-art deep convolutional neural network. Our results show that SNNs with LIF neurons can provide competitive denoising performance but at a reduced computational cost.