Madry A, Makelov A, Schmidt L, et al. Towards Deep Learning Models Resistant to Adversarial Attacks.[J]. arXiv: Machine Learning, 2017.

@article{madry2017towards,

title={Towards Deep Learning Models Resistant to Adversarial Attacks.},

author={Madry, Aleksander and Makelov, Aleksandar and Schmidt, Ludwig and Tsipras, Dimitris and Vladu, Adrian},

journal={arXiv: Machine Learning},

year={2017}}

概

利用特定的方法产生"坏"样本(Adversarial samples), 以此来促进网络的稳定性是当下的热点之一, 本文以实验为主, 比较PGD( projected gradient descent) 和 FGSM(fast gradient sign method)在不同数据下的表现, 以及由普通样本产生"坏"样本会出现的一些现象.

主要内容

Adversarial attacks 主要聚焦于下列问题:

\[\tag{2.1}
\min_{\theta} \rho (\theta) \quad where \quad \rho(\theta) =\mathbb{E}_{(x,y)\sim D}[\max_{\delta \in S} L(\theta, x+\delta, y)].
\]

其中\(S\)是我们指定的摄动集合, 直接一点就是\(|\delta| <constant\)之类.

通过FGSM产生"坏"样本:

\[x + \epsilon \: \mathrm{sgn}(\nabla_x L(\theta,x,y)).
\]

这个思想是很直接的(从线性感知器谈起, 具体看here).

PGD的思路是, 给定摄动集\(S\), 比如小于某个常数的摄动(e.g. \(\{\tilde{x}:\|x-\tilde{x}\|_{\infty}<c\}\)), 多次迭代寻找合适的adversarial samples:

\[x^{t+1} = \prod_{x+S} (x^t + \alpha \: \mathrm{sgn} (\nabla_x L(\theta,x, y)),
\]

其中\(\prod\)表示投影算子, 假设\(S=\{\tilde{x}:\|x-\tilde{x}\|_{\infty}<c\}\),

\[x^{t+1} = \arg \min_{z \in x+S} \frac{1}{2} \|z - (x^t + \alpha \: \mathrm{sgn} (\nabla_x L(\theta,x, y))\|_2^2,
\]

实际上, 可以分开讨论第\((i,j)\)个元素, \(y:=(x^t + \alpha \: \mathrm{sgn} (\nabla_x L(\theta,x, y))\), 只需找到\(z_{ij}\)使得

\[\|z_{ij}-y_{ij}\|_2
\]

最小即可. 此时有显示解为:

\[z_{ij}=
\left \{
\begin{array}{ll}
x_{ij} +c & y_{ij} > x_{ij}+c \\
x_{ij} -c & y_{ij} < x_{ij}-c \\
y_{ij} & else.
\end{array} \right.
\]

简而言之就是一个截断.

重复几次, 至到\(x^t\)被判断的类别与初始的\(x\)不同或者达到最大迭代次数.

Note

  • 如果我们训练网络能够免疫PGD的攻击, 那么其也能很大一部分其它的攻击.
  • FGSM对抗训练不能提高网络的稳定性(在摄动较大的时候).
  • weak models may fail to learn non-trival classfiers.
  • 网络越强(参数等程度)训练出来的稳定性越好, 同时可转移(指adversarial samples 在多个网络中被误判)会变差.

Towards Deep Learning Models Resistant to Adversarial Attacks的更多相关文章

  1. How to Grid Search Hyperparameters for Deep Learning Models in Python With Keras

    Hyperparameter optimization is a big part of deep learning. The reason is that neural networks are n ...

  2. a Javascript library for training Deep Learning models

    w强化算法和数学,来迎接机器学习.神经网络. http://cs.stanford.edu/people/karpathy/convnetjs/ ConvNetJS is a Javascript l ...

  3. Run Your Tensorflow Deep Learning Models on Google AI

    People commonly tend to put much effort on hyperparameter tuning and training while using Tensoflow& ...

  4. What are some good books/papers for learning deep learning?

    What's the most effective way to get started with deep learning?       29 Answers     Yoshua Bengio, ...

  5. (转) Awesome Deep Learning

    Awesome Deep Learning  Table of Contents Free Online Books Courses Videos and Lectures Papers Tutori ...

  6. (转)分布式深度学习系统构建 简介 Distributed Deep Learning

    HOME ABOUT CONTACT SUBSCRIBE VIA RSS   DEEP LEARNING FOR ENTERPRISE Distributed Deep Learning, Part ...

  7. The Brain vs Deep Learning Part I: Computational Complexity — Or Why the Singularity Is Nowhere Near

    The Brain vs Deep Learning Part I: Computational Complexity — Or Why the Singularity Is Nowhere Near ...

  8. Paper Reading——LEMNA:Explaining Deep Learning based Security Applications

    Motivation: The lack of transparency of the deep  learning models creates key barriers to establishi ...

  9. Coursera Deep Learning 2 Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization - week1, Assignment(Regularization)

    声明:所有内容来自coursera,作为个人学习笔记记录在这里. Regularization Welcome to the second assignment of this week. Deep ...

随机推荐

  1. 学习java 7.26

    学习内容: 进度条是图形界面中广浅个较大的文件时,操作系统会显示一个进度条,用于标识复制操作完成的比例:当启动Eclipse等程序时,因为需要加载较多的资源,故而启动速度较慢,程序也会在启动过程中显示 ...

  2. ache

    ache和pain可能没啥差别,头疼和头好痛都对.从词典来看,有backache, bellyache, earache, headache, heartache, moustache/mustach ...

  3. 内存管理——new delete expression

    C++申请释放内存的方法与详情表 调用情况 1.new expression new表达式在申请内存过程中都发生了什么? 编译器将new这个分解为下面的主要3步代码,①首先调用operator new ...

  4. C++11的auto自动推导类型

    auto是C++11的类型推导关键字,很强大 例程看一下它的用法 #include<vector> #include<algorithm> #include<functi ...

  5. IntentFilter,PendingIntent

    1.当Intent在组件间传递时,组件如果想告知Android系统自己能够响应那些Intent,那么就需要用到IntentFilter对象. IntentFilter对象负责过滤掉组件无法响应和处理的 ...

  6. virtualBox 系统移植

    把virtualbox已经存在的系统移植到其他机器. 1.把系统如下文件考到一个安装了virtualbox的机器. 2.点击控制-->注册 然后浏览到复制的文件路径. 3.修改uuid 不管是l ...

  7. 【Linux】【Shell】【text】sed

        sed [OPTION]...  'script'  [input-file] ...         script:             地址定界编辑命令                 ...

  8. Linux下部署Java项目(jetty作为容器)常用脚本命令

    startup.sh #!/bin/bash echo $(basename $(pwd)) "jetty started" cd jetty nohup java -Xmx8g ...

  9. vue cli3.0 首次加载优化

    项目经理要求做首页加载优化,打包后从十几兆优化到两兆多,记下来怕下次忘记 运行report脚本 可看到都加载了那些内容,在从dist文件中index.html 查看首次加载都加载了那些东西,如下图:然 ...

  10. Linux 文件属性及详细操作

    目录 Linux 文件属性 文件属性信息组成 文件属性概念说明 文件软硬链接说明 硬链接 软链接 补充知识 存储数据相关-inode/block inode: block: 读取文件原理图 Linux ...