arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2409.05294cs.CRcs.AIcs.LG
收藏

TERD:保护扩散模型免受后门攻击的统一框架

TERD: A Unified Framework for Safeguarding Diffusion Models Against Backdoors

Peking University(北京大学);CISPA Helmholtz Center for Information Security(CISPA亥姆霍兹信息安全研究中心)

Yichuan Mo, Hui Huang, Mingjie Li, Ang Li, Yisen Wang

AI总结:

本文提出TERD统一后门防御框架,通过逆向触发器恢复、噪声空间输入检测及KL散度模型检测,在多分辨率数据集上实现100% TPR和TNR。

摘要:

扩散模型在图像生成领域取得了显著成功,但它们仍极易受到后门攻击:当出现预定义触发器时,模型会生成特定的不良输出,从而破坏其完整性。本文研究如何保护扩散模型免受这一危险威胁。具体而言,我们提出TERD,这是一种后门防御框架,对当前攻击进行统一建模,使我们能够推导出可获得的逆向损失。进一步采用触发器还原策略:先利用从先验分布采样的噪声获得触发器的初始近似,再通过差分多步采样器进行精化。此外,借助还原的触发器,我们从噪声空间进行后门检测,提出首个面向扩散模型的后门输入检测方法,以及一种新颖的模型检测算法,该算法计算逆向分布与良性分布之间的KL散度。大量评估表明,TERD在不同分辨率的数据集上均获得100%的真正率(TPR)和真负率(TNR)。TERD还对其他基于随机微分方程(SDE)的模型表现出良好的适应性。代码位于https://github.com/PKU-ML/TERD。

英文摘要:

Diffusion models have achieved notable success in image generation, but they remain highly vulnerable to backdoor attacks, which compromise their integrity by producing specific undesirable outputs when presented with a pre-defined trigger. In this paper, we investigate how to protect diffusion models from this dangerous threat. Specifically, we propose TERD, a backdoor defense framework that builds unified modeling for current attacks, which enables us to derive an accessible reversed loss. A trigger reversion strategy is further employed: an initial approximation of the trigger through noise sampled from a prior distribution, followed by refinement through differential multi-step samplers. Additionally, with the reversed trigger, we propose backdoor detection from the noise space, introducing the first backdoor input detection approach for diffusion models and a novel model detection algorithm that calculates the KL divergence between reversed and benign distributions. Extensive evaluations demonstrate that TERD secures a 100% True Positive Rate (TPR) and True Negative Rate (TNR) across datasets of varying resolutions. TERD also demonstrates nice adaptability to other Stochastic Differential Equation (SDE)-based models. Our code is available at https://github.com/PKU-ML/TERD.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑