mlsanityData to 7 Oct 2026

Bridged Adversarial Training

arXiv 2108.11135 ↗First posted 25 Aug 2021cs.LG cs.AIcited by 3

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 47 of its 61 references that are in the map. First posted on arXiv 25 Aug 2021.

Builds on

47 of 61 references.

References · 47

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  3. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  4. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  5. 2013Intriguing properties of neural networkscited by 5,114
  6. 2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 4,993
  7. 2016Wide Residual Networkscited by 3,057
  8. 2016Towards Evaluating the Robustness of Neural Networkscited by 2,976
  9. 2016Identity Mappings in Deep Residual Networkscited by 1,960
  10. 2019Theoretically Principled Trade-off between Robustness and Accuracycited by 1,510
  11. 2020Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attackscited by 1,442
  12. 2020Sharpness-Aware Minimization for Efficiently Improving Generalizationcited by 1,404
35 more
  1. 2018Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examplescited by 1,259
  2. 2018Averaging Weights Leads to Wider Optima and Better Generalizationcited by 1,194
  3. 2017Ensemble Adversarial Training: Attacks and Defensescited by 833
  4. 2019Square Attack: a query-efficient black-box adversarial attack via random searchcited by 682
  5. 2015Cyclical Learning Rates for Training Neural Networkscited by 627
  6. 2020On Adaptive Attacks to Adversarial Example Defensescited by 547
  7. 2020Overfitting in adversarially robust deep learningcited by 504
  8. 2019Unlabeled Data Improves Adversarial Robustnesscited by 415
  9. 2019Minimally distorted Adversarial Examples with a Fast Adaptive Boundary Attackcited by 333
  10. 2019Provably Robust Deep Learning via Adversarially Trained Smoothed Classifierscited by 328
  11. 2019Efficient and Accurate Estimation of Lipschitz Constants for Deep Neural Networkscited by 290
  12. 2017Parseval Networks: Improving Robustness to Adversarial Examplescited by 288
  13. 2020Uncovering the Limits of Adversarial Training against Norm-Bounded Adversarial Examplescited by 275
  14. 2017Improving the Adversarial Robustness and Interpretability of Deep Neural Networks by Regularizing their Input Gradientscited by 236
  15. 2018Adversarial Logit Pairingcited by 224
  16. 2018Sorting out Lipschitz function approximationcited by 211
  17. 2021Fixing Data Augmentation to Improve Adversarial Robustnesscited by 201
  18. 2020Bag of Tricks for Adversarial Trainingcited by 199
  19. 2017Formal Guarantees on the Robustness of a Classifier against Adversarial Manipulationcited by 182
  20. 2018MMA Training: Direct Input Space Margin Maximization through Adversarial Trainingcited by 171
  21. 2019Are Labels Required for Improving Adversarial Robustness?cited by 162
  22. 2020Torchattacks: A PyTorch Repository for Adversarial Attackscited by 156
  23. 2016Robust Large Margin Deep Neural Networkscited by 117
  24. 2018Large Margin Deep Networks for Classificationcited by 110
  25. 2019Adversarially Robust Generalization Just Requires More Unlabeled Datacited by 75
  26. 2020Understanding Catastrophic Overfitting in Single-step Adversarial Trainingcited by 75
  27. 2020On the Loss Landscape of Adversarial Training: Identifying Challenges and How to Overcome Themcited by 59
  28. 2021Probabilistic Margins for Instance Reweighting in Adversarial Trainingcited by 57
  29. 2021Exploring Memorization in Adversarial Trainingcited by 55
  30. 2019Robustness to Adversarial Perturbations in Learning from Incomplete Datacited by 49
  31. 2020How benign is benign overfitting?cited by 40
  32. 2020Efficient Robust Training via Backward Smoothingcited by 29
  33. 2020Boundary thickness and robustness in learning modelscited by 26
  34. 2020Using Single-Step Adversarial Training to Defend Iterative Adversarial Examplescited by 9
  35. 2020Towards Deep Learning Models Resistant to Large Perturbationscited by 8

Built on by

Cited by year

Citing papers per year: 2022: 2; 2023: 1.

3 papers.

  1. 2022Removing Batch Normalization Boosts Adversarial Trainingcited by 24
  2. 2022Squeeze Training for Adversarial Robustnesscited by 10
  3. 2023Interpolation for Robust Learning: Data Augmentation on Wasserstein Geodesicscited by 3