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 2309.00007cs.CVcs.CRcs.LG
收藏

当度量不可靠时:针对Top-$k$多标签学习的不可察觉对抗扰动

When Measures are Unreliable: Imperceptible Adversarial Perturbations toward Top-$k$ Multi-Label Learning

Peng Cheng Laboratory(鹏城实验室)

Yuchen Sun, Qianqian Xu, Zitai Wang, Qingming Huang

AI总结:

本文提出度量不可察觉性概念,设计新损失函数与高效凸优化算法,生成同时欺骗视觉与逃避度量监控的对抗扰动,在多个大型基准上验证了攻击top-$k$多标签系统的优越性。

摘要:

随着深度神经网络的巨大成功,对抗学习在从多类学习到多标签学习的各种研究中受到了广泛关注。然而,现有的针对多标签学习的对抗攻击只追求传统的视觉不可察觉性,而忽略了来自诸如Precision@$k$和mAP@$k$等度量的新的可感知问题。具体来说,当一个训练良好的多标签分类器在某些样本上表现远低于预期时,受害者很容易意识到这种性能退化源于攻击,而不是模型本身。因此,理想的多标签对抗攻击不仅应设法欺骗视觉感知,还应逃避度量的监控。为此,本文首先提出了度量不可察觉性的概念。然后,设计了一种新颖的损失函数来生成能够同时实现视觉和度量不可察觉性的对抗扰动。此外,建立了一种高效的算法,该算法具有凸目标,以优化此目标。最后,在大型基准数据集(如PASCAL VOC 2012、MS COCO和NUS WIDE)上的广泛实验证明了我们提出的方法在攻击top-$k$多标签系统方面的优越性。

英文摘要:

With the great success of deep neural networks, adversarial learning has received widespread attention in various studies, ranging from multi-class learning to multi-label learning. However, existing adversarial attacks toward multi-label learning only pursue the traditional visual imperceptibility but ignore the new perceptible problem coming from measures such as Precision@$k$ and mAP@$k$. Specifically, when a well-trained multi-label classifier performs far below the expectation on some samples, the victim can easily realize that this performance degeneration stems from attack, rather than the model itself. Therefore, an ideal multi-labeling adversarial attack should manage to not only deceive visual perception but also evade monitoring of measures. To this end, this paper first proposes the concept of measure imperceptibility. Then, a novel loss function is devised to generate such adversarial perturbations that could achieve both visual and measure imperceptibility. Furthermore, an efficient algorithm, which enjoys a convex objective, is established to optimize this objective. Finally, extensive experiments on large-scale benchmark datasets, such as PASCAL VOC 2012, MS COCO, and NUS WIDE, demonstrate the superiority of our proposed method in attacking the top-$k$ multi-label systems.

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。

↑