arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2408.14728cs.LGcs.AIcs.CR
收藏

通过对抗训练的切线空间视角提升清洁准确性

Improving Clean Accuracy via a Tangent-Space Perspective on Adversarial Training

Department of Statistics and Operations Research, University of North Carolina at Chapel Hill(统计与运筹学系,北卡罗来纳大学教堂山分校);Department of Mathematics, Purdue University(数学系,普渡大学)

Bongsoo Yi, Rongjie Lai, Yao Li

AI总结:

本文提出TART方法,通过利用数据流形几何提升清洁数据准确性,同时保持对抗鲁棒性。

摘要:

对抗训练已被证明能提高深度神经网络对对抗攻击的鲁棒性。然而,这种增强的鲁棒性往往以清洁数据准确性的大幅下降为代价。在本文中,我们通过引入切线方向引导对抗训练(TART)方法来解决这一限制。我们主张具有大法向分量的对抗示例会过度扭曲决策边界并降低清洁准确性。TART通过估计对抗示例的切线方向,并根据其切向分量的范数自适应地调节扰动边界来解决这一问题。据我们所知,TART是首个将切线空间和方向概念明确纳入对抗训练的对抗防御框架。在合成和基准数据集上的广泛实验表明,TART在保持对抗鲁棒性的同时,能够一致地提升清洁准确性。

英文摘要:

Adversarial training has proven effective in improving the robustness of deep neural networks against adversarial attacks. However, this enhanced robustness often comes at the cost of a substantial drop in accuracy on clean data. In this paper, we address this limitation by introducing Tangent Direction Guided Adversarial Training (TART), a novel method that enhances clean accuracy by exploiting the geometry of the data manifold. We argue that adversarial examples with large components in the normal direction can overly distort the decision boundary and degrade clean accuracy. TART addresses this issue by estimating the tangent direction of adversarial examples and adaptively modulating the perturbation bound based on the norm of their tangential component. To the best of our knowledge, TART is the first adversarial defense framework that explicitly incorporates the concept of tangent space and direction into adversarial training. Extensive experiments on both synthetic and benchmark datasets demonstrate that TART consistently improves clean accuracy while maintaining robustness against adversarial attacks.

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。

↑