深度学习
Sadvine
这个作者很懒,什么都没留下…
展开
专栏收录文章
- 默认排序
- 最新发布
- 最早发布
- 最多阅读
- 最少阅读
-
【论文回顾】Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Exa
Paper notes:Obfuscated Gradients include three unprotected gradients:Shattered Gradients,Stochastic Gradients,Exploding & Vanishing Gradients5 behaivor to show defense with these wrong gr...原创 2020-01-17 16:36:44 · 1321 阅读 · 0 评论 -
【论文回顾】Prior Convictions: Black-Box Adversarial Attacks with Bandits and Priors
Paper notes:this paperfocused on black-box and untargeted threat model.the attack problem:pre-requsite: loss funtion L(x,y), distance metric l-p norm.non-convexity but well solved by previo...原创 2020-01-15 15:05:45 · 1245 阅读 · 4 评论 -
【论文回顾】Towards Evaluating the Robustness of Neural Networks
paper notes:1.this paper introduced three new attacks for L-0 L-2 L-infin distance metrics by defining different choices of objective function. Specifically, they are all based on the L-2 attack.2...原创 2020-01-11 20:17:20 · 811 阅读 · 0 评论 -
【论文回顾】The Limitations of Deep Learning in Adversarial Settings
The Limitations of Deep Learning in Adversarial Settingspaper notes:This paper introduces the background of adversarial examples including adversarial goals, capabilities and then explains how to ...原创 2020-01-11 15:31:27 · 568 阅读 · 0 评论
分享