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 2006.07258cs.LGstat.ML
收藏

D-square-B:类自然对抗攻击的深度分布界

D-square-B: Deep Distribution Bound for Natural-looking Adversarial Attack

Qiuling Xu, Guanhong Tao, Xiangyu Zhang

摘要:

我们提出了一种新的技术,可以通过分布分位数界限和多项式障碍损失函数来限制某些深层中内部激活值引起的变化,从而生成看起来自然的对抗性示例。通过绑定模型内部而不是单个像素,我们的攻击承认与原始输入的现有特征密切相关的扰动,允许生成的示例看起来自然,同时与原始输入具有不同且通常相当大的像素距离。强制执行每个神经元分布分位数界限允许解决内部激活值的不均匀性。我们对ImageNet和五种不同模型架构的评估表明,我们的攻击非常有效。与最先进的像素空间攻击,语义攻击和特征空间攻击相比,我们的攻击可以实现相同的攻击成功/置信水平,同时具有更自然的对抗性扰动。这些扰动依赖于现有的局部特征,并且没有任何固定的像素边界。

英文摘要:

We propose a novel technique that can generate natural-looking adversarial examples by bounding the variations induced for internal activation values in some deep layer(s), through a distribution quantile bound and a polynomial barrier loss function. By bounding model internals instead of individual pixels, our attack admits perturbations closely coupled with the existing features of the original input, allowing the generated examples to be natural-looking while having diverse and often substantial pixel distances from the original input. Enforcing per-neuron distribution quantile bounds allows addressing the non-uniformity of internal activation values. Our evaluation on ImageNet and five different model architecture demonstrates that our attack is quite effective. Compared to the state-of-the-art pixel space attack, semantic attack, and feature space attack, our attack can achieve the same attack success/confidence level while having much more natural-looking adversarial perturbations. These perturbations piggy-back on existing local features and do not have any fixed pixel bounds.

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。

↑