Data Free Backdoor Attacks

TL;DR

This paper introduces DFBA, a novel data-free, retraining-free backdoor attack method that modifies a few classifier parameters to inject undetectable backdoors, achieving high success rates and bypassing defenses.

Contribution

DFBA is the first approach to inject backdoors without retraining or data, maintaining model architecture and ensuring stealthiness and effectiveness.

Findings

Achieves 100% attack success rate

Bypasses six state-of-the-art defenses

Incurs negligible classification loss

Abstract

Backdoor attacks aim to inject a backdoor into a classifier such that it predicts any input with an attacker-chosen backdoor trigger as an attacker-chosen target class. Existing backdoor attacks require either retraining the classifier with some clean data or modifying the model's architecture. As a result, they are 1) not applicable when clean data is unavailable, 2) less efficient when the model is large, and 3) less stealthy due to architecture changes. In this work, we propose DFBA, a novel retraining-free and data-free backdoor attack without changing the model architecture. Technically, our proposed method modifies a few parameters of a classifier to inject a backdoor. Through theoretical analysis, we verify that our injected backdoor is provably undetectable and unremovable by various state-of-the-art defenses under mild assumptions. Our evaluation on multiple datasets further demonstrates that our injected backdoor: 1) incurs negligible classification loss, 2) achieves 100% attack success rates, and 3) bypasses six existing state-of-the-art defenses. Moreover, our comparison with a state-of-the-art non-data-free backdoor attack shows our attack is more stealthy and effective against various defenses while achieving less classification accuracy loss.