DetectX -- Adversarial Input Detection using Current Signatures in Memristive XBar Arrays

TL;DR

DetectX is a hardware-efficient adversarial input detection method using memristive crossbar signatures, significantly improving energy efficiency and robustness against various attacks through a dual-phase training process.

Contribution

This work introduces DetectX, a novel hardware-friendly adversarial detection mechanism leveraging memristive crossbar signatures and a dual-phase training methodology for enhanced robustness.

Findings

DetectX achieves 10x-25x energy efficiency over prior methods.

It attains ROC-AUC > 0.95 for strong white-box and black-box attacks.

DetectX demonstrates robustness against dynamic adversarial attacks.

Abstract

Adversarial input detection has emerged as a prominent technique to harden Deep Neural Networks(DNNs) against adversarial attacks. Most prior works use neural network-based detectors or complex statistical analysis for adversarial detection. These approaches are computationally intensive and vulnerable to adversarial attacks. To this end, we propose DetectX - a hardware friendly adversarial detection mechanism using hardware signatures like Sum of column Currents (SoI) in memristive crossbars (XBar). We show that adversarial inputs have higher SoI compared to clean inputs. However, the difference is too small for reliable adversarial detection. Hence, we propose a dual-phase training methodology: Phase1 training is geared towards increasing the separation between clean and adversarial SoIs; Phase2 training improves the overall robustness against different strengths of adversarial attacks. For hardware-based adversarial detection, we implement the DetectX module using 32nm CMOS circuits and integrate it with a Neurosim-like analog crossbar architecture. We perform hardware evaluation of the Neurosim+DetectX system on the Neurosim platform using datasets-CIFAR10(VGG8), CIFAR100(VGG16) and TinyImagenet(ResNet18). Our experiments show that DetectX is 10x-25x more energy efficient and immune to dynamic adversarial attacks compared to previous state-of-the-art works. Moreover, we achieve high detection performance (ROC-AUC > 0.95) for strong white-box and black-box attacks. The code has been released at https://github.com/Intelligent-Computing-Lab-Yale/DetectX