Efficient Adversarial Training with Transferable Adversarial Examples
Preprint
Abstract
Adversarial training is an effective defense method to protect classification models against adversarial attacks. However, one limitation of this approach is that it can require orders of magnitude additional training time due to high cost of generating strong adversarial examples during training. In this paper, we first show that there is high transferability between models from neighboring epochs in the same training process, i.e., adversarial examples from one epoch continue to be adversarial in subsequent epochs. Leveraging this property, we propose a novel method, Adversarial Training with Transferable Adversarial Examples (ATTA), that can enhance the robustness of trained models and greatly improve the training efficiency by accumulating adversarial perturbations through epochs. Compared to state-of-the-art adversarial training methods, ATTA enhances adversarial accuracy by up to 7.2% on CIFAR10 and requires 12~14x less training time on MNIST and CIFAR10 datasets with comparable model robustness.Keywords
All Related Versions
This publication has 6 references indexed in Scilit:
- Universal Adversarial TrainingProceedings of the AAAI Conference on Artificial Intelligence, 2020
- Learning Transferable Adversarial Examples via Ghost NetworksProceedings of the AAAI Conference on Artificial Intelligence, 2020
- A Discussion of 'Adversarial Examples Are Not Bugs, They Are Features'Distill, 2019
- Learning to Attack: Adversarial Transformation NetworksProceedings of the AAAI Conference on Artificial Intelligence, 2018
- Wide Residual Networks2016
- Gradient-based learning applied to document recognitionProceedings of the IEEE, 1998