Reconstructive Neuron Pruning for Backdoor Defense

TL;DR

This paper introduces Reconstructive Neuron Pruning (RNP), a novel method that unlearns and recovers neurons to effectively expose and prune backdoor neurons in neural networks, improving defense against backdoor attacks.

Contribution

The paper proposes a new asymmetric reconstructive learning method for backdoor defense that unlearns and recovers neurons to effectively remove backdoor effects.

Findings

RNP achieves state-of-the-art backdoor defense performance.

Unlearned models can enhance other backdoor detection tasks.

Effective pruning of backdoor neurons across various attack types.

Abstract

Deep neural networks (DNNs) have been found to be vulnerable to backdoor attacks, raising security concerns about their deployment in mission-critical applications. While existing defense methods have demonstrated promising results, it is still not clear how to effectively remove backdoor-associated neurons in backdoored DNNs. In this paper, we propose a novel defense called \emph{Reconstructive Neuron Pruning} (RNP) to expose and prune backdoor neurons via an unlearning and then recovering process. Specifically, RNP first unlearns the neurons by maximizing the model's error on a small subset of clean samples and then recovers the neurons by minimizing the model's error on the same data. In RNP, unlearning is operated at the neuron level while recovering is operated at the filter level, forming an asymmetric reconstructive learning procedure. We show that such an asymmetric process on only a few clean samples can effectively expose and prune the backdoor neurons implanted by a wide range of attacks, achieving a new state-of-the-art defense performance. Moreover, the unlearned model at the intermediate step of our RNP can be directly used to improve other backdoor defense tasks including backdoor removal, trigger recovery, backdoor label detection, and backdoor sample detection. Code is available at \url{https://github.com/bboylyg/RNP}.