arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2008.08847cs.CV
收藏

又是中级攻击

Yet Another Intermediate-Level Attack

ByteDance AI Lab(字节跳动人工智能实验室);University of California, Davis(加利福尼亚大学戴维斯分校)

Qizhang Li, Yiwen Guo, Hao Chen

摘要:

对抗性示例在深度神经网络(DNN)模型中的可移植性是一系列黑盒攻击的关键。在本文中,我们提出了一种新的方法来增强基线对抗示例的黑盒可移植性。通过建立中间水平差异的线性映射(一组对抗性输入和它们的良性对应物之间)来预测诱发的对抗性损失,我们的目标是充分利用多步基线攻击的优化过程。我们在CIFAR-100和ImageNet上进行了大量的实验来验证我们的方法的有效性。实验结果表明,它大大优于以前的国家的最先进的。我们的代码在https://github.com/qizhangli/ila-plus-plus。

英文摘要:

The transferability of adversarial examples across deep neural network (DNN) models is the crux of a spectrum of black-box attacks. In this paper, we propose a novel method to enhance the black-box transferability of baseline adversarial examples. By establishing a linear mapping of the intermediate-level discrepancies (between a set of adversarial inputs and their benign counterparts) for predicting the evoked adversarial loss, we aim to take full advantage of the optimization procedure of multi-step baseline attacks. We conducted extensive experiments to verify the effectiveness of our method on CIFAR-100 and ImageNet. Experimental results demonstrate that it outperforms previous state-of-the-arts considerably. Our code is at https://github.com/qizhangli/ila-plus-plus.

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。

↑