Hardware Accelerator for Adversarial Attacks on Deep Learning Neural Networks

TL;DR

This paper introduces a hardware accelerator based on memristor crossbar arrays to enhance the efficiency of adversarial attack algorithms on deep neural networks, aiming to improve the robustness and security of future AI systems.

Contribution

It presents the first hardware accelerator for adversarial attacks on DNNs, significantly boosting throughput and efficiency, and proposes four implementations tailored to different performance needs.

Findings

Significant throughput improvement in adversarial attack systems

Enhanced energy and computational efficiency of the accelerator

Potential to strengthen DNN robustness and security

Abstract

Recent studies identify that Deep learning Neural Networks (DNNs) are vulnerable to subtle perturbations, which are not perceptible to human visual system but can fool the DNN models and lead to wrong outputs. A class of adversarial attack network algorithms has been proposed to generate robust physical perturbations under different circumstances. These algorithms are the first efforts to move forward secure deep learning by providing an avenue to train future defense networks, however, the intrinsic complexity of them prevents their broader usage. In this paper, we propose the first hardware accelerator for adversarial attacks based on memristor crossbar arrays. Our design significantly improves the throughput of a visual adversarial perturbation system, which can further improve the robustness and security of future deep learning systems. Based on the algorithm uniqueness, we propose four implementations for the adversarial attack accelerator ($A^3$) to improve the throughput, energy efficiency, and computational efficiency.