Bridged Adversarial Training
arXiv:2108.11135
Abstract
Adversarial robustness is considered as a required property of deep neural networks. In this study, we discover that adversarially trained models might have significantly different characteristics in terms of margin and smoothness, even they show similar robustness. Inspired by the observation, we investigate the effect of different regularizers and discover the negative effect of the smoothness regularizer on maximizing the margin. Based on the analyses, we propose a new method called bridged adversarial training that mitigates the negative effect by bridging the gap between clean and adversarial examples. We provide theoretical and empirical evidence that the proposed method provides stable and better robustness, especially for large perturbations.
References in corpus (26)
- PyTorch: An Imperative Style, High-Performance Deep Learning Library
- SGDR: Stochastic Gradient Descent with Warm Restarts
- Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examples
- Theoretically Principled Trade-off between Robustness and Accuracy
- Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks
- Robustness and Regularization of Support Vector Machines
- Improving the Adversarial Robustness and Interpretability of Deep Neural Networks by Regularizing their Input Gradients
- Formal Guarantees on the Robustness of a Classifier against Adversarial Manipulation
- Provably Robust Deep Learning via Adversarially Trained Smoothed Classifiers
- Robust Large Margin Deep Neural Networks
- Are Labels Required for Improving Adversarial Robustness?
- Uncovering the Limits of Adversarial Training against Norm-Bounded Adversarial Examples
- On Adaptive Attacks to Adversarial Example Defenses
- Minimally distorted Adversarial Examples with a Fast Adaptive Boundary Attack
- Fixing Data Augmentation to Improve Adversarial Robustness
- Torchattacks: A PyTorch Repository for Adversarial Attacks
- MMA Training: Direct Input Space Margin Maximization through Adversarial Training
- Adversarially Robust Generalization Just Requires More Unlabeled Data
- Efficient and Accurate Estimation of Lipschitz Constants for Deep Neural Networks
- Overfitting in adversarially robust deep learning
- On the Loss Landscape of Adversarial Training: Identifying Challenges and How to Overcome Them
- Exploring Memorization in Adversarial Training
- Large Margin Deep Networks for Classification
- Bag of Tricks for Adversarial Training
- How benign is benign overfitting?
- Towards Deep Learning Models Resistant to Large Perturbations