arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2004.00225cs.LGcs.AIcs.CRcs.CVstat.ML
收藏

MetaPoison:实用的通用清洁标签数据投毒

MetaPoison: Practical General-purpose Clean-label Data Poisoning

University of Maryland(马里兰大学);University of Siegen(锡根大学);United States Naval Academy(美国海军学院)

W. Ronny Huang, Jonas Geiping, Liam Fowl, Gavin Taylor, Tom Goldstein

摘要:

数据投毒-攻击者通过对训练数据的子集进行不可察觉的更改来控制模型的过程-是神经网络环境中的一种新兴威胁。针对数据投毒神经网络的现有攻击依赖于手工制作的算法,因为直接通过双层优化来解决中毒问题通常被认为是深度模型难以处理的。我们提出了MetaPoison,这是一种一阶方法,通过元学习和工艺毒药来近似双层问题,从而欺骗神经网络。MetaPoison是有效的:它比以前的清洁标签中毒方法有很大的优势。MetaPoison是健壮的:为一个模型制作的中毒数据可以传输到具有未知训练设置和架构的各种受害者模型。MetaPoison是通用的,它不仅适用于微调场景,还适用于从头开始的端到端训练,到目前为止,这对于使用深度网络的清洁标签攻击来说是不可行的。MetaPoison可以实现任意的对手目标--比如使用一个类别的毒药来使目标图像贴上另一个任意选择的类别的标签。最后,MetaPoison在现实世界中发挥作用。我们首次展示了在黑盒Google Cloud AutoML API上训练的模型的成功数据投毒。代码和预制毒药提供在https://github.com/wronnyhuang/metapoison

英文摘要:

Data poisoning -- the process by which an attacker takes control of a model by making imperceptible changes to a subset of the training data -- is an emerging threat in the context of neural networks. Existing attacks for data poisoning neural networks have relied on hand-crafted heuristics, because solving the poisoning problem directly via bilevel optimization is generally thought of as intractable for deep models. We propose MetaPoison, a first-order method that approximates the bilevel problem via meta-learning and crafts poisons that fool neural networks. MetaPoison is effective: it outperforms previous clean-label poisoning methods by a large margin. MetaPoison is robust: poisoned data made for one model transfer to a variety of victim models with unknown training settings and architectures. MetaPoison is general-purpose, it works not only in fine-tuning scenarios, but also for end-to-end training from scratch, which till now hasn't been feasible for clean-label attacks with deep nets. MetaPoison can achieve arbitrary adversary goals -- like using poisons of one class to make a target image don the label of another arbitrarily chosen class. Finally, MetaPoison works in the real-world. We demonstrate for the first time successful data poisoning of models trained on the black-box Google Cloud AutoML API. Code and premade poisons are provided at https://github.com/wronnyhuang/metapoison

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。

↑