Privacy-Preserving Discretized Spiking Neural Networks

TL;DR

This paper introduces a privacy-preserving framework for discretized spiking neural networks (SNNs) using homomorphic encryption, achieving high accuracy on encrypted data while maintaining efficiency and low energy consumption.

Contribution

The paper proposes FHE-DiSNN, a novel framework enabling homomorphic evaluation of SNNs with privacy preservation and improved computational efficiency.

Findings

Achieves 94.4% accuracy on MNIST with encrypted data.

Maintains high accuracy with only 0.6% decrease compared to original SNN.

Demonstrates superiority of SNN over second-generation neural networks for privacy-preserving evaluation.

Abstract

The rapid development of artificial intelligence has brought considerable convenience, yet also introduces significant security risks. One of the research hotspots is to balance data privacy and utility in the real world of artificial intelligence. The present second-generation artificial neural networks have made tremendous advances, but some big models could have really high computational costs. The third-generation neural network, SNN (Spiking Neural Network), mimics real neurons by using discrete spike signals, whose sequences exhibit strong sparsity, providing advantages such as low energy consumption and high efficiency. In this paper, we construct a framework to evaluate the homomorphic computation of SNN named FHE-DiSNN that enables SNN to achieve good prediction performance on encrypted data. First, benefitting from the discrete nature of spike signals, our proposed model avoids the errors introduced by discretizing activation functions. Second, by applying bootstrapping, we design new private preserving functions FHE-Fire and FHE-Reset, through which noise can be refreshed, allowing us to evaluate SNN for an arbitrary number of operations. Furthermore, We improve the computational efficiency of FHE-DiSNN while maintaining a high level of accuracy. Finally, we evaluate our model on the MNIST dataset. The experiments show that FHE-DiSNN with 30 neurons in the hidden layer achieves a minimum prediction accuracy of 94.4%. Under optimal parameters, it achieves a 95.1% accuracy, with only a 0.6% decrease compared to the original SNN (95.7%). These results demonstrate the superiority of SNN over second-generation neural networks for homomorphic evaluation.