CAT: Customized Adversarial Training for Improved Robustness

Authors: Minhao Cheng, Qi Lei, Pin-Yu Chen, Inderjit Dhillon, Cho-Jui Hsieh

IJCAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental through extensive experiments, we show that the proposed algorithm achieves better clean and robust accuracy than previous adversarial training methods.
Researcher Affiliation Collaboration 1Department of Computer Science and Engineering, HKUST 2Department of Electrical and Computer Engineering, Princeton University 3Department of Computer Science, UT Austin 4Department of Computer Science, UCLA 5IBM Research AI 6Amazon
Pseudocode Yes Algorithm 1 CAT algorithm
Open Source Code Yes Our code is publicly available at https://github.com/cmhcbb/CAT-Customized-Adversarial-Training-for-Improved-Robustness.
Open Datasets Yes We use CIFAR-10 dataset for performance evaluation.
Dataset Splits No The paper mentions using CIFAR-10 and standard models from other works, implying standard splits, but does not explicitly state the training, validation, and test dataset splits (e.g., percentages or counts) within the text.
Hardware Specification Yes All our experiments were implemented in Pytorch-1.4 and conducted using a GTX 2080 TI GPU.
Software Dependencies Yes All our experiments were implemented in Pytorch-1.4
Experiment Setup Yes We set the number of iterations in adversarial attack to be 10 for all methods during training. Adversarial training and TRADES are trained on PGD attacks setting ϵ = 8/255 with cross entropy loss (CE). All the models are trained using SGD with momentum 0.9, weight decay 5 10 4. For VGG-16/Wide Res Net models, we use the initial learning rate of 0.01/0.1, and we decay the learning rate by 90% at the 80th, 140th, and 180th epoch. For CAT, we set epsilon scheduling parameter η = 0.005, ϵmax = 8/255 and weighting parameter c = 10. We set β = 1 for the distribution Dirichlet(β), which is equal to a uniform distribution. Also, we set κ = 10.