JARN_ICLR2020

This is our Tensorflow implementation of Jacobian Adversarially Regularized Networks (JARN).

Jacobian Adversarially Regularized Networks for Robustness (ICLR 2020)
Alvin Chan, Yi Tay, Yew Soon Ong, Jie Fu
https://arxiv.org/abs/1912.10185

TL;DR: We show that training classifiers to produce salient input Jacobian matrices with a GAN-like regularization can boost adversarial robustness.

Dependencies

1. Tensorflow 1.14.0
2. Python 3.7

Usage

1. Install dependencies with pip install -r requirements.txt.
2. Run JARN training and evaluation with sh run_train_jarn.sh. Final evaluation output is saved in attack_log.

Code overview

• train_jarn.py: trains the JARN model and subsequently evaluate on adversarial examples.
• pgd_attack.py: generates adversarial examples and save them in attacks/.
• run_attack.py: evaluates model on adversarial examples from attacks/.
• config.py: training parameters for JARN.
• config_attack.py: parameters for adversarial example evaluation.
• model_jarn.py: contains code for JARN model architectures.
• cifar10_input.py provides utility functions and classes for loading the CIFAR10 dataset.

Citation

If you find our repository useful, please consider citing our paper:

@article{chan2019jacobian,
  title={Jacobian Adversarially Regularized Networks for Robustness},
  author={Chan, Alvin and Tay, Yi and Ong, Yew Soon and Fu, Jie},
  journal={arXiv preprint arXiv:1912.10185},
  year={2019}
}

Acknowledgements

Useful code bases we used in our work:

• https://github.com/MadryLab/cifar10_challenge (for adversarial example generation and evaluation)