BriefGPT.xyz
May, 2019
对抗训练的规范不可知鲁棒性
On Norm-Agnostic Robustness of Adversarial Training
HTML
PDF
Bai Li, Changyou Chen, Wenlin Wang, Lawrence Carin
TL;DR
本文研究对抗样本攻击机器学习模型并提出一种新的攻击方法,证明最先进的对抗训练方法无法同时获得对$\ell_2$和$\ell_\infty$范数的健壮性,提出可能的解决方案及其局限性。
Abstract
adversarial examples
are carefully perturbed in-puts for fooling machine learning models. A well-acknowledged defense method against such examples is
adversarial training
, where
→