Quarantine: Sparsity Can Uncover the Trojan Attack Trigger for Free

TL;DR

This paper introduces a novel method for detecting Trojan attacks in deep neural networks by leveraging network sparsity, identifying a sparse subnetwork that retains Trojan features while being benign on clean inputs.

Contribution

The paper proposes a new Trojan detection approach based on network pruning and the lottery ticket hypothesis, effective even without clean training data.

Findings

Effective detection across multiple datasets and architectures

Trojan features are more stable to pruning than benign features

Achieves high accuracy in isolating Trojan information

Abstract

Trojan attacks threaten deep neural networks (DNNs) by poisoning them to behave normally on most samples, yet to produce manipulated results for inputs attached with a particular trigger. Several works attempt to detect whether a given DNN has been injected with a specific trigger during the training. In a parallel line of research, the lottery ticket hypothesis reveals the existence of sparse subnetworks which are capable of reaching competitive performance as the dense network after independent training. Connecting these two dots, we investigate the problem of Trojan DNN detection from the brand new lens of sparsity, even when no clean training data is available. Our crucial observation is that the Trojan features are significantly more stable to network pruning than benign features. Leveraging that, we propose a novel Trojan network detection regime: first locating a "winning Trojan lottery ticket" which preserves nearly full Trojan information yet only chance-level performance on clean inputs; then recovering the trigger embedded in this already isolated subnetwork. Extensive experiments on various datasets, i.e., CIFAR-10, CIFAR-100, and ImageNet, with different network architectures, i.e., VGG-16, ResNet-18, ResNet-20s, and DenseNet-100 demonstrate the effectiveness of our proposal. Codes are available at https://github.com/VITA-Group/Backdoor-LTH.