Magnitude-based Neuron Pruning for Backdoor Defens

TL;DR

This paper introduces a magnitude-based neuron pruning method that effectively detects and removes backdoor neurons in deep neural networks, enhancing security with limited clean data.

Contribution

It proposes a novel magnitude-guided pruning approach that exploits the deviation in neuron magnitude-saliency correlation to defend against backdoor attacks.

Findings

Achieves state-of-the-art backdoor defense performance

Effectively detects backdoor neurons with limited clean data

Preserves clean model performance after pruning

Abstract

Deep Neural Networks (DNNs) are known to be vulnerable to backdoor attacks, posing concerning threats to their reliable deployment. Recent research reveals that backdoors can be erased from infected DNNs by pruning a specific group of neurons, while how to effectively identify and remove these backdoor-associated neurons remains an open challenge. In this paper, we investigate the correlation between backdoor behavior and neuron magnitude, and find that backdoor neurons deviate from the magnitude-saliency correlation of the model. The deviation inspires us to propose a Magnitude-based Neuron Pruning (MNP) method to detect and prune backdoor neurons. Specifically, MNP uses three magnitude-guided objective functions to manipulate the magnitude-saliency correlation of backdoor neurons, thus achieving the purpose of exposing backdoor behavior, eliminating backdoor neurons and preserving clean neurons, respectively. Experiments show our pruning strategy achieves state-of-the-art backdoor defense performance against a variety of backdoor attacks with a limited amount of clean data, demonstrating the crucial role of magnitude for guiding backdoor defenses.