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 1607.02533cs.CVcs.CRcs.LGstat.ML
收藏

物理世界中的对抗性例子

Adversarial examples in the physical world

Google Brain(谷歌大脑);OpenAI

Alexey Kurakin, Ian Goodfellow, Samy Bengio

摘要:

大多数现有的机器学习分类器非常容易受到对抗性样本的影响。一个对抗性的例子是输入数据的样本被轻微地修改了,这是为了导致机器学习分类器错误地分类它。在许多情况下,这些修改是如此的微妙,以至于人类观察者甚至根本没有注意到修改,但分类器仍然会出错。对抗性示例会带来安全问题,因为它们可能被用来对机器学习系统进行攻击,即使对手无法访问底层模型。到目前为止,所有以前的工作都假设了一个威胁模型,其中对手可以将数据直接输入机器学习分类器。对于在物理世界中运行的系统,例如使用来自摄像机和其他传感器的信号作为输入的系统,情况并非总是如此。本文表明,即使在这样的物理世界场景中,机器学习系统也容易受到对抗性示例的影响。我们通过将从手机摄像头获得的对抗图像馈送到ImageNet Inception分类器并测量系统的分类精度来证明这一点。我们发现,即使通过摄像头感知,也有很大一部分对抗性样本被错误分类。

英文摘要:

Most existing machine learning classifiers are highly vulnerable to adversarial examples. An adversarial example is a sample of input data which has been modified very slightly in a way that is intended to cause a machine learning classifier to misclassify it. In many cases, these modifications can be so subtle that a human observer does not even notice the modification at all, yet the classifier still makes a mistake. Adversarial examples pose security concerns because they could be used to perform an attack on machine learning systems, even if the adversary has no access to the underlying model. Up to now, all previous work have assumed a threat model in which the adversary can feed data directly into the machine learning classifier. This is not always the case for systems operating in the physical world, for example those which are using signals from cameras and other sensors as an input. This paper shows that even in such physical world scenarios, machine learning systems are vulnerable to adversarial examples. We demonstrate this by feeding adversarial images obtained from cell-phone camera to an ImageNet Inception classifier and measuring the classification accuracy of the system. We find that a large fraction of adversarial examples are classified incorrectly even when perceived through the camera.

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。

↑