Free Adversarial Training

This repository belongs to the Free Adversarial Training paper. The implementation is inspired by CIFAR10 Adversarial Example Challenge so to them we give the credit. This repo is for the CIFAR-10 and CIFAR-100 datasets and is in Tensorflow. Our Free-m models can acheive comparable performance with conventional PGD adversarial training at a fraction of the time.

News!: We have released our ImageNet implementation of Free adversarial training in Pytorch !

CIFAR-10 WRN 32-10 (L-inf epsilon=8):

Model	Natural	PGD-100	CW-100	10 restart PGD-20	train-time (min)
Natrual	95.01	0.00	0.00	0.00	780
Free-2	91.45	33.20	34.57	33.41	816
Free-4	87.83	40.35	41.96	40.73	800
Free-8	85.96	46.19	46.60	46.33	785
Free-10	83.94	45.79	45.86	45.94	785
Madry 7-PGD (public model)	87.25	45.29	46.52	45.53	5418

CIFAR-100 WRN 32-10 (L-inf epsilon=8):

Model	Natural	PGD-20	PGD-100	train-time (min)
Natrual	78.84	0.00	0.00	811
Free-2	69.20	15.37	14.86	816
Free-4	65.28	20.64	20.15	767
Free-8	62.13	25.88	25.58	780
Free-10	59.27	25.15	24.88	776
Madry 2-PGD trained	67.94	17.08	16.50	2053
Madry 7-PGD trained	59.87	22.76	22.52	5157

Demo

To train a new robust model for free! run the following command specifying the replay parameter m:

python free_train.py -m 8

To evaluate a robust model using PGD-20 with 2 random restarts run:

python multi_restart_pgd_attack.py --model_dir $MODEL_DIR --num_restarts 2

Note that if you have trained a CIFAR-100 model, even for evaluation, you should pass the dataset argument. For example:

python multi_restart_pgd_attack.py --model_dir $MODEL_DIR_TO_CIFAR100 --num_restarts 2 -d cifar100

Requirements

To install all the requirements plus tensorflow for multi-gpus run: (Inspired By Illarion ikhlestov )

pip install -r requirements/gpu.txt

Alternatively, to install the requirements plus tensorflow for cpu run:

pip install -r requirements/cpu.txt

To prepare the data, please see Datasets section.

If you find the paper or the code useful for your study, please consider citing the free training paper:

@article{shafahi2019adversarial,
  title={Adversarial Training for Free!},
  author={Shafahi, Ali and Najibi, Mahyar and Ghiasi, Amin and Xu, Zheng and Dickerson, John and Studer, Christoph and Davis, Larry S and Taylor, Gavin and Goldstein, Tom},
  journal={arXiv preprint arXiv:1904.12843},
  year={2019}
}