BriefGPT.xyz
Mar, 2018
使用基函数变换防御对抗性图像
Defending against Adversarial Images using Basis Functions Transformations
HTML
PDF
Uri Shaham, James Garritano, Yutaro Yamada, Ethan Weinberger, Alex Cloninger...
TL;DR
本文研究了通过对图像的基函数表示进行操作来防御深度学习中的对抗攻击的效果，发现JPEG压缩和低通滤波有较高的防御效果，并提出了一种新的基于白盒攻击的数学推导方法。
Abstract
We study the effectiveness of various approaches that defend against
adversarial attacks
on
deep networks
via manipulations based on basis function representations of images. Specifically, we experiment with
→