README.md

Awesome CV Adversarial Attack List

This repository is a curated list of papers and open source code about competition for CV Adversarial Attack.

Paper

• Adversarial Examples in Modern Machine Learning: A Review [Paper]
• Adversarial Attacks and Defenses in Images, Graphs and Text: A Review [Paper]
• Adversarial Atacks and Defences: A Survey [Paper]
• Adversarial Examples: Opportunities and Challenges [Paper]
• Threat of Adversarial Attacks on Deep Learning in Computer Vision: A Survey [Paper]
• Adversarial Examples: Attacks and Defenses for Deep Learning [Paper]
• Adversarial Attacks and Defences Competition [Paper]

Competition and Solution

• NIPS 2017: Targeted Adversarial Attack
  dongyp13 : 1th place [code] [Boosting Adversarial Attacks with Momentum]
  rwightman : 5th place[code]
  rwightman : pytorch-nips2017-attack-example [code]
  anlthms [code]
  erko [code]
  Stanford&Suns [code]
  ysharma1126 [code]
  EdwardTyantov [code]
  ckomaki [code]

• NIPS 2017: Non-Targeted Adversarial Attack
  dongyp13 : 1th place [code]
  toshi-k : 5th place [code]

• NIPS 2017: DefenseAgainst Adversarial Attack
  cihangxie : 2th place [code]
  lfz [code]
  Roy-YL [code]

• CAAD2018
  ysharma1126 [Tech Report]
  FAIR & JHU team [Tech Report]
  TSAIL team [Attack Tech Report]
  TSAIL team [Defense Tech Report]
  DLight team [Tech Report]
  NorthWest Sec team [Tech Report]
  Teaflow team [Tech Report]
  RNG team [Tech Report]
  Kunlin team [Tech Report]
  Official source code [code]
  jxjrework [code]
  tianweiy [code]
  geekpwn [code]
  0three [code]

• MCS 2018. Adversarial Attacks on Black Box Face Recognition
  snakers4 [code] [Tech Report]

• IJCAI-19: 阿里巴巴人工智能对抗算法竞赛
  Jowekk : 2th place in the defense track [code]
  jiangyangzhou : 5th place in the non-targted attack track [code]
  cshanjiewu : 6th place in the targted attack track [code]
  yfreedomliTHU [code]

• 天池：安全AI挑战者计划第一期-人脸识别对抗
  BruceQFWang : 4th place [code]
  tabsun : 10th place [code]
  InoriJam : 11th place [code]
  SunnyWangGitHub [code]

Repository

• cleverhans [Link]
• foolbox [Link]
• adversarial-robustness-toolbox (ART) [Link]
• Adversarial-Examples-Reading-List [Link]
• adversarial-example-pytorch [Link]
• advertorch [Link]
• EvadeML-Zoo [Link]
• U-Turn [Link]

Tutorials

• 知乎：如何看待机器视觉的“对抗样本”问题，其原理是什么？
• 对抗样本
• NIPS 2017对抗样本攻防竞赛总结 [Link]
• 产生和防御对抗样本的新方法|分享总结 [Link]
• 清华参赛队攻击组论文 [Boosting Adversarial Attacks with Momentum]
• 清华参赛队防御组论文 [Defense against Adversarial Attacks Using High-Level Representation Guided Denoiser]
• 动量迭代攻击和高层引导去噪：对抗样本攻防的新方法 [Video]
• AI安全大佬教你如何攻击云端图像分类模型|纯干货 [Link]
• IJCV2022 | 逆转特征让re-id模型从88.54%到0.15% [Link]