https://github.com/guanghelee/neurips19-certificates-of-robustness

"Tight Certificates of Adversarial Robustness for Randomly Smoothed Classifiers" (NeurIPS 2019, previously called "A Stratified Approach to Robustness for Randomly Smoothed Classifiers")
https://github.com/guanghelee/neurips19-certificates-of-robustness

Last synced: 3 months ago
JSON representation

"Tight Certificates of Adversarial Robustness for Randomly Smoothed Classifiers" (NeurIPS 2019, previously called "A Stratified Approach to Robustness for Randomly Smoothed Classifiers")

• Host: GitHub
• URL: https://github.com/guanghelee/neurips19-certificates-of-robustness
• Owner: guanghelee
• Created: 2019-07-01T18:11:39.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2019-11-16T04:06:50.000Z (almost 5 years ago)
• Last Synced: 2024-05-18T22:23:18.440Z (6 months ago)
• Language: Python
• Homepage:
• Size: 6.53 MB
• Stars: 17
• Watchers: 3
• Forks: 1
• Open Issues: 1
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome-decision-tree-papers - [Code

README

        
# Tight Certificates of Adversarial Robustness for Randomly Smoothed Classifiers:

This repository is for the paper

 * "[Tight Certificates of Adversarial Robustness for Randomly Smoothed Classifiers](https://arxiv.org/pdf/1906.04948.pdf)" by [Guang-He Lee](https://people.csail.mit.edu/guanghe/), [Yang Yuan](http://www.callowbird.com), [Shiyu Chang](http://people.csail.mit.edu/chang87/), and [Tommi S. Jaakkola](http://people.csail.mit.edu/tommi/) in NeurIPS 2019.

 * The old title for this paper is "A Stratified Approach to Robustness for Randomly Smoothed Classifiers"

## Outline

 

 * Please see each experiment in the corresponding directory (and the README therein).

 * The MNIST experiment has been released. 

 * The ImageNet experiment has been released. (Not carefully checked. Please let me know if you find any problem.)

 * The pre-computed ρ-1r(0.5) and trained ResNet50 models have been released for the ImageNet experiment.

 * If you want to compute your own ρ-1r(0.5), please see the examples in the MNIST or ImageNet folder.

 * Please let me know ([email protected]) if you need the codes for the decision tree experiment.

## Citation:

If you find this repo useful for your research, please cite the paper

```

@inproceedings{lee2019tight,

  title={Tight Certificates of Adversarial Robustness for Randomly Smoothed Classifiers},

  author={Guang-He Lee and Yang Yuan and Shiyu Chang and Tommi S. Jaakkola},

  booktitle={Advances in Neural Information Processing Systems},

  year={2019}

}

```