arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2409.13945cs.AI
收藏

PureDiffusion:在生成式扩散模型中使用后门对抗后门

PureDiffusion: Using Backdoor to Counter Backdoor in Generative Diffusion Models

INRS, University of Québec(魁北克大学INRS)

Vu Tuan Truong, Long Bao Le

AI总结:

本文提出PureDiffusion,一种通过反转后门触发器来高效检测扩散模型后门攻击的防御框架,在保真度和后门成功率上显著优于现有方法。

摘要:

扩散模型(DMs)是先进的深度学习模型,在广泛的生成任务上取得了最先进的能力。然而,最近的研究表明它们在后门攻击方面存在脆弱性,当模型输入包含后门触发器时,被植入后门的DM会持续生成一个指定结果(例如有害图像),称为后门目标。尽管已经研究了各种后门技术来攻击DM,但针对这些威胁的防御方法仍然有限且研究不足,特别是在反转后门触发器方面。在本文中,我们介绍了PureDiffusion,这是一种新颖的后门防御框架,能够通过反转嵌入在DM中的后门触发器来高效检测后门攻击。我们在各种触发器-目标对上进行了广泛实验,结果表明PureDiffusion在保真度(即反转的触发器与原始触发器的相似程度)和后门成功率(即反转的触发器导致相应后门目标的比率)方面,以巨大差距优于现有防御方法。值得注意的是,在某些情况下,PureDiffusion反转的后门触发器甚至比原始触发器实现了更高的攻击成功率。

英文摘要:

Diffusion models (DMs) are advanced deep learning models that achieved state-of-the-art capability on a wide range of generative tasks. However, recent studies have shown their vulnerability regarding backdoor attacks, in which backdoored DMs consistently generate a designated result (e.g., a harmful image) called backdoor target when the models' input contains a backdoor trigger. Although various backdoor techniques have been investigated to attack DMs, defense methods against these threats are still limited and underexplored, especially in inverting the backdoor trigger. In this paper, we introduce PureDiffusion, a novel backdoor defense framework that can efficiently detect backdoor attacks by inverting backdoor triggers embedded in DMs. Our extensive experiments on various trigger-target pairs show that PureDiffusion outperforms existing defense methods with a large gap in terms of fidelity (i.e., how much the inverted trigger resembles the original trigger) and backdoor success rate (i.e., the rate that the inverted trigger leads to the corresponding backdoor target). Notably, in certain cases, backdoor triggers inverted by PureDiffusion even achieve higher attack success rate than the original triggers.

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。

↑