Adversarial Training with Rectified Rejection
The code for the paper Adversarial Training with Rectified Rejection.
Environment settings and libraries we used in our experiments
This project is tested under the following environment settings:
- OS: Ubuntu 18.04.4
- GPU: Geforce 2080 Ti or Tesla P100
- Cuda: 10.1, Cudnn: v7.6
- Python: 3.6
- PyTorch: >= 1.6.0
- Torchvision: >= 0.6.0
Acknowledgement
The codes are modifed based on Rice et al. 2020, and the model architectures are implemented by pytorch-cifar.
Training Commands
Below we provide running commands training the models with the RR module, taking the setting of PGD-AT + RR (ResNet-18) as an example:
python train_cifar.py --model_name PreActResNet18_twobranch_DenseV1 --attack pgd --lr-schedule piecewise \
--epochs 110 --epsilon 8 \
--attack-iters 10 --pgd-alpha 2 \
--fname auto \
--batch-size 128 \
--adaptivetrain --adaptivetrainlambda 1.0 \
--weight_decay 5e-4 \
--twobranch --useBN \
--selfreweightCalibrate \
--dataset 'CIFAR-10' \
--ATframework 'PGDAT' \
--SGconfidenceW
The FLAG --model_name can be PreActResNet18_twobranch_DenseV1 (ResNet-18) or WideResNet_twobranch_DenseV1 (WRN-34-10). For alternating different AT frameworks, we can set the FLAG --ATframework to be one of PGDAT, TRADES, CCAT.
Evaluation Commands
Below we provide running commands for evaluations.
Evaluating under the PGD attacks
The trained model is saved at trained_models/model_path, where the specific name of model_path is automatically generated during training. The command for evaluating under PGD attacks is:
python eval_cifar.py --model_name PreActResNet18_twobranch_DenseV1 --evalset test --norm l_inf --epsilon 8 \
--attack-iters 1000 --pgd-alpha 2 \
--fname trained_models/model_path \
--load_epoch -1 \
--dataset 'CIFAR-10' \
--twobranch --useBN \
--selfreweightCalibrate
Evaluating under the adaptive CW attacks
The parameter FLAGs --binary_search_steps, --CW_iter, --CW_confidence can be changed, where --detectmetric indicates the rejector that needs to be adaptively evaded.
python eval_cifar_CW.py --model_name PreActResNet18_twobranch_DenseV1 --evalset adaptiveCWtest \
--fname trained_models/model_path \
--load_epoch -1 --seed 2020 \
--binary_search_steps 9 --CW_iter 100 --CW_confidence 0 \
--threatmodel linf --reportmodel linf \
--twobranch --useBN \
--selfreweightCalibrate \
--detectmetric 'RR' \
--dataset 'CIFAR-10'
Evaluating under multi-target and GAMA attacks
The running command for evaluating under multi-target attacks is activated by the FLAG --evalonMultitarget as:
python eval_cifar.py --model_name PreActResNet18_twobranch_DenseV1 --evalset test --norm l_inf --epsilon 8 \
--attack-iters 100 --pgd-alpha 2 \
--fname trained_models/model_path \
--load_epoch -1 \
--dataset 'CIFAR-10' \
--twobranch --useBN \
--selfreweightCalibrate \
--evalonMultitarget --restarts 1
The running command for evaluating under GAMA attacks is activated by the FLAG --evalonGAMA_PGD or --evalonGAMA_FW as:
python eval_cifar.py --model_name PreActResNet18_twobranch_DenseV1 --evalset test --norm l_inf --epsilon 8 \
--attack-iters 100 --pgd-alpha 2 \
--fname trained_models/model_path \
--load_epoch -1 \
--dataset 'CIFAR-10' \
--twobranch --useBN \
--selfreweightCalibrate \
--evalonGAMA_FW
Evaluating under CIFAR-10-C
The running command for evaluating on common corruptions in CIFAR-10-C is:
python eval_cifar_CIFAR10-C.py --model_name PreActResNet18_twobranch_DenseV1 \
--fname trained_models/model_path \
--load_epoch -1 \
--dataset 'CIFAR-10' \
--twobranch --useBN \
--selfreweightCalibrate
62 Nov 23, 2022
56 Dec 22, 2022
70 Oct 29, 2022
74 Dec 03, 2022
34 Mar 31, 2022
90 Jan 06, 2023
318 Dec 20, 2022
3.5k Jan 08, 2023
0 Aug 04, 2021
11 Nov 07, 2022
1 Jan 12, 2022
194 Dec 27, 2022
269 Jan 02, 2023
2.2k Jan 09, 2023
134 Jul 10, 2022
33 Dec 06, 2022
14 Nov 22, 2022
282 Dec 28, 2022
2 Dec 27, 2022
62 Dec 21, 2022