Adversarial-Divergence-Reduction (ADR)

This implementation corresponds with the paper "Improving Adversarial Robustness by Enforcing Local and Global Compactness" which has been accepted to ECCV-2020. paperslide

We provide either tensorflow (original version) and pytorch (recently added) implementations.

Update log

• June 2021, porting to pytorch.
• July 2020, release the tensorflow implementation.

Cite

If you find this work useful for your research, please consider citing:

@inproceedings{bui2020improving,
  title={Improving Adversarial Robustness by Enforcing Local and Global Compactness},
  author={Bui, Anh and Le, Trung and Zhao, He and Montague, Paul and deVel, Olivier and Abraham, Tamas and Phung, Dinh},
  booktitle={European Conference on Computer Vision (ECCV), 2020},
  year={2020}
}

or

@inproceedings{DBLP:conf/eccv/BuiLZMDAP20,
  author    = {Anh Bui and
               Trung Le and
               He Zhao and
               Paul Montague and
               Olivier Y. DeVel and
               Tamas Abraham and
               Dinh Q. Phung},
  editor    = {Andrea Vedaldi and
               Horst Bischof and
               Thomas Brox and
               Jan{-}Michael Frahm},
  title     = {Improving Adversarial Robustness by Enforcing Local and Global Compactness},
  booktitle = {Computer Vision - {ECCV} 2020 - 16th European Conference, Glasgow,
               UK, August 23-28, 2020, Proceedings, Part {XXVII}},
  series    = {Lecture Notes in Computer Science},
  volume    = {12372},
  pages     = {209--223},
  publisher = {Springer},
  year      = {2020},
  url       = {https://doi.org/10.1007/978-3-030-58583-9\_13},
  doi       = {10.1007/978-3-030-58583-9\_13},
  timestamp = {Tue, 01 Dec 2020 09:11:51 +0100},
  biburl    = {https://dblp.org/rec/conf/eccv/BuiLZMDAP20.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

License

As a free open-source implementation, ADR is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. All other warranties including, but not limited to, merchantability and fitness for purpose, whether express, implied, or arising by operation of law, course of dealing, or trade usage are hereby disclaimed. I believe that the programs compute what I claim they compute, but I do not guarantee this. The programs may be poorly and inconsistently documented and may contain undocumented components, features or modifications. I make no guarantee that these programs will be suitable for any application.

ADR is licensed under the Apache License 2.0.