HoSNN: Adversarially-Robust Homeostatic Spiking Neural Networks with Adaptive Firing Thresholds

TL;DR

This paper introduces HoSNN, a robust spiking neural network model utilizing adaptive firing thresholds inspired by neural homeostasis, significantly enhancing adversarial robustness across multiple datasets.

Contribution

The paper proposes a novel TA-LIF neuron model with dynamic thresholds, improving robustness of SNNs against adversarial attacks through a biologically-inspired homeostatic mechanism.

Findings

HoSNN achieves up to 74.91% accuracy on FashionMNIST under adversarial attack.

Theoretical analysis confirms stability and error suppression of TA-LIF neurons.

Significant robustness improvements over standard LIF-based SNNs across datasets.

Abstract

While spiking neural networks (SNNs) offer a promising neurally-inspired model of computation, they are vulnerable to adversarial attacks. We present the first study that draws inspiration from neural homeostasis to design a threshold-adapting leaky integrate-and-fire (TA-LIF) neuron model and utilize TA-LIF neurons to construct the adversarially robust homeostatic SNNs (HoSNNs) for improved robustness. The TA-LIF model incorporates a self-stabilizing dynamic thresholding mechanism, offering a local feedback control solution to the minimization of each neuron's membrane potential error caused by adversarial disturbance. Theoretical analysis demonstrates favorable dynamic properties of TA-LIF neurons in terms of the bounded-input bounded-output stability and suppressed time growth of membrane potential error, underscoring their superior robustness compared with the standard LIF neurons. When trained with weak FGSM attacks (attack budget = 2/255) and tested with much stronger PGD attacks (attack budget = 8/255), our HoSNNs significantly improve model accuracy on several datasets: from 30.54% to 74.91% on FashionMNIST, from 0.44% to 35.06% on SVHN, from 0.56% to 42.63% on CIFAR10, from 0.04% to 16.66% on CIFAR100, over the conventional LIF-based SNNs.