Calibrated Adversarial Training
This addresses a challenge in robust machine learning for security-critical applications, but appears incremental as it builds on existing adversarial training methods.
The paper tackles the problem of adversarial training's adverse effects from semantic perturbations in adversarial examples, and presents Calibrated Adversarial Training, which shows superior performance on multiple public datasets.
Adversarial training is an approach of increasing the robustness of models to adversarial attacks by including adversarial examples in the training set. One major challenge of producing adversarial examples is to contain sufficient perturbation in the example to flip the model's output while not making severe changes in the example's semantical content. Exuberant change in the semantical content could also change the true label of the example. Adding such examples to the training set results in adverse effects. In this paper, we present the Calibrated Adversarial Training, a method that reduces the adverse effects of semantic perturbations in adversarial training. The method produces pixel-level adaptations to the perturbations based on novel calibrated robust error. We provide theoretical analysis on the calibrated robust error and derive an upper bound for it. Our empirical results show a superior performance of the Calibrated Adversarial Training over a number of public datasets.