ADBM: Adversarial diffusion bridge model for reliable adversarial purification

Autor:	Li, Xiao, Sun, Wenxuan, Chen, Huanran, Li, Qiongxiu, Liu, Yining, He, Yingzhe, Shi, Jie, Hu, Xiaolin
Rok vydání:	2024
Předmět:	Computer Science - Machine Learning Computer Science - Artificial Intelligence Computer Science - Computer Vision and Pattern Recognition
Druh dokumentu:	Working Paper
Popis:	Recently Diffusion-based Purification (DiffPure) has been recognized as an effective defense method against adversarial examples. However, we find DiffPure which directly employs the original pre-trained diffusion models for adversarial purification, to be suboptimal. This is due to an inherent trade-off between noise purification performance and data recovery quality. Additionally, the reliability of existing evaluations for DiffPure is questionable, as they rely on weak adaptive attacks. In this work, we propose a novel Adversarial Diffusion Bridge Model, termed ADBM. ADBM directly constructs a reverse bridge from the diffused adversarial data back to its original clean examples, enhancing the purification capabilities of the original diffusion models. Through theoretical analysis and experimental validation across various scenarios, ADBM has proven to be a superior and robust defense mechanism, offering significant promise for practical applications. Comment: 20 pages
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2408.00315 Zobrazit plný text záznamu View this record from Arxiv