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 2211.14860cs.CVcs.AI
收藏

挫败深度神经网络中的解释

Foiling Explanations in Deep Neural Networks

Ben-Gurion University(本-古里安大学);DeepKeep

Snir Vitrack Tamam, Raz Lapid, Moshe Sipper

AI总结:

本文提出黑盒、模型无关的 AttaXAI 攻击,仅利用输出 logits 与解释图,通过进化策略施加不可察觉扰动以任意控制 XAI 解释图。

摘要:

在过去十年中,深度神经网络(DNN)已深刻影响众多领域。然而,尽管其在许多问题上表现卓越,黑盒特性仍给可解释性带来重大挑战。事实上,可解释人工智能(XAI)在若干领域至关重要;在这些领域中,仅有答案而不说明该答案如何得出,几乎没有价值。本文揭示了面向图像的 DNN 解释方法的一个令人不安的性质:通过对输入图像进行微小视觉改变——几乎不影响网络输出——我们展示了如何利用进化策略任意操纵解释。我们的新算法 AttaXAI 是一种针对 XAI 算法的模型无关对抗攻击,只需要访问分类器的输出 logits 和解释图;这些弱假设使我们的方法在涉及真实世界模型和数据时极具实用性。我们在两个基准数据集 CIFAR100 和 ImageNet 上,使用四种不同的预训练深度学习模型比较方法性能:VGG16-CIFAR100、VGG16-ImageNet、MobileNet-CIFAR100 和 Inception-v3-ImageNet。我们发现,无需使用梯度或其他模型内部信息,即可操纵 XAI 方法。我们的新算法能够成功以人眼不可察觉的方式操纵图像,使 XAI 方法输出特定解释图。据我们所知,这是黑盒环境中的首个此类方法;我们认为,在希望、需要或法律强制要求可解释性的场景中,它具有重要价值。

英文摘要:

Deep neural networks (DNNs) have greatly impacted numerous fields over the past decade. Yet despite exhibiting superb performance over many problems, their black-box nature still poses a significant challenge with respect to explainability. Indeed, explainable artificial intelligence (XAI) is crucial in several fields, wherein the answer alone -- sans a reasoning of how said answer was derived -- is of little value. This paper uncovers a troubling property of explanation methods for image-based DNNs: by making small visual changes to the input image -- hardly influencing the network's output -- we demonstrate how explanations may be arbitrarily manipulated through the use of evolution strategies. Our novel algorithm, AttaXAI, a model-agnostic, adversarial attack on XAI algorithms, only requires access to the output logits of a classifier and to the explanation map; these weak assumptions render our approach highly useful where real-world models and data are concerned. We compare our method's performance on two benchmark datasets -- CIFAR100 and ImageNet -- using four different pretrained deep-learning models: VGG16-CIFAR100, VGG16-ImageNet, MobileNet-CIFAR100, and Inception-v3-ImageNet. We find that the XAI methods can be manipulated without the use of gradients or other model internals. Our novel algorithm is successfully able to manipulate an image in a manner imperceptible to the human eye, such that the XAI method outputs a specific explanation map. To our knowledge, this is the first such method in a black-box setting, and we believe it has significant value where explainability is desired, required, or legally mandatory.

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。

↑