AdvRush
Official Code for AdvRush: Searching for Adversarially Robust Neural Architectures (ICCV '21)
Environmental Set-up
Python == 3.6.12, PyTorch == 1.2.0, torchvision == 0.4.0
AdvRush Search Process
cd advrush && python train_search.py --batch_size 32 --gpu 0 --epochs 60 --a_gamma 0.01 --a_warmup_epochs 0 --w_warmup_epochs 60 --loss_hessian loss_cure
Adversarial Training
cd advrush && python adv_train.py --batch_size 64 --gpu 0 --epochs 200 --adv_loss pgd --arch ADVRUSH
Evaluation under PGD Attack
Prior to the evaluation process, add all necessary checkpoint files (preferably in the form of .pth.tar) to the /eval/checkpoints folder. To conduct white-box attacks,
cd eval &&
python pgd_attack.py --white-box-attack True --test-batch-size 10 --arch [arch_name] --checkpoint [./checkpoints/file_name.pth.tar] --data_type [cifar10/svhn]
To conduct black-box attacks,
cd eval &&
python pgd_attack.py --test-batch-size 10 --target_arch [target_arch] --target_checkpoint [./checkpoints/target_file.pth.tar] --source_arch [source_arch] --source_checkpoint [./checkpoints/source_file.pth.tar] --data_type cifar10
References
DARTS: Differentiable Architecture Search [ICLR '19] code paper
Robustness via Curvature Regularization, and Vice Versa [CVPR '19] code paper
Tradeoff-inspired Adversarial Defense via Surrogate-loss Minimization [ICML '19] code paper
57 Nov 14, 2022
590 Dec 22, 2022
1 Sep 19, 2022
29 Dec 07, 2022
6 May 15, 2022
36 Dec 30, 2022
14.5k Jan 08, 2023
1.2k Jan 07, 2023
325 Oct 20, 2022
1.3k Dec 23, 2022
13 Nov 03, 2022
109 Dec 16, 2022
3 Sep 21, 2022
22 Mar 13, 2022
27 Dec 27, 2022
45 Nov 17, 2022
2 Oct 04, 2022
49 Nov 21, 2022
97 Dec 23, 2022
1.7k Jan 06, 2023