BriefGPT.xyz
May, 2019
使用对抗网络进行鲁棒性深度学习的直接方法
A Direct Approach to Robust Deep Learning Using Adversarial Networks
HTML
PDF
Huaxia Wang, Chun-Nam Yu
TL;DR
该论文提出了一种基于生成对抗网络(GAN)框架下的新防御机制来对抗黑盒攻击,在经验上表现良好并能与利用梯度下降的集成对抗训练和对抗训练等最先进的方法媲美。
Abstract
Deep
neural networks
have been shown to perform well in many classical machine learning problems, especially in image classification tasks. However, researchers have found that
neural networks
can be easily foole
→