mlsanityData to 7 Oct 2026

Rethinking the Effect of Data Augmentation in Adversarial Contrastive Learning

arXiv 2303.01289 ↗First posted 2 Mar 2023cs.LGcited by 20

Not on the map. Why?

Abstract (arXiv)

Recent works have shown that self-supervised learning can achieve remarkable robustness when integrated with adversarial training (AT). However, the robustness gap between supervised AT (sup-AT) and self-supervised AT (self-AT) remains significant. Motivated by this observation, we revisit existing self-AT methods and discover an inherent dilemma that affects self-AT robustness: either strong or weak data augmentations are harmful to self-AT, and a medium strength is insufficient to bridge the gap. To resolve this dilemma, we propose a simple remedy named DYNACL (Dynamic Adversarial Contrastive Learning). In particular, we propose an augmentation schedule that gradually anneals from a strong augmentation to a weak one to benefit from both extreme cases. Besides, we adopt a fast post-processing stage for adapting it to downstream tasks. Through extensive experiments, we show that DYNACL can improve state-of-the-art self-AT robustness by 8.84% under Auto-Attack on the CIFAR-10 dataset, and can even outperform vanilla supervised adversarial training for the first time. Our code is available at urlhttps://github.com/PKU-ML/DYNACL.

Key facts. Cited by 20 papers in the map (data through 7 Oct 2026). Builds on 34 of its 40 references that are in the map. First posted on arXiv 2 Mar 2023.

Builds on

34 of 40 references.

References · 34

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  3. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
  4. 2018Representation Learning with Contrastive Predictive Codingcited by 7,662
  5. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
  6. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  7. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  8. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,712
  9. 2021An Empirical Study of Training Self-Supervised Vision Transformerscited by 1,610
  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. 2020Big Self-Supervised Models are Strong Semi-Supervised Learnerscited by 1,375
22 more
  1. 2019Contrastive Multiview Codingcited by 1,297
  2. 2017Visualizing the Loss Landscape of Neural Netscited by 1,126
  3. 2022Fine-Tuning can Distort Pretrained Features and Underperform Out-of-Distributioncited by 722
  4. 2020Overfitting in adversarially robust deep learningcited by 504
  5. 2017Certifying Some Distributional Robustness with Principled Adversarial Trainingcited by 333
  6. 2021Provable Guarantees for Self-Supervised Deep Learning with Spectral Contrastive Losscited by 311
  7. 2019Adversarial Examples Improve Image Recognitioncited by 283
  8. 2020Bag of Tricks for Adversarial Trainingcited by 199
  9. 2019Are Labels Required for Improving Adversarial Robustness?cited by 162
  10. 2020Robust Pre-Training by Adversarial Contrastive Learningcited by 144
  11. 2020Adversarial Self-Supervised Contrastive Learningcited by 138
  12. 2021On the Convergence and Robustness of Adversarial Trainingcited by 127
  13. 2021Solo-learn: A Library of Self-supervised Methods for Visual Representation Learningcited by 116
  14. 2022Chaos is a Ladder: A New Theoretical Understanding of Contrastive Learning via Augmentation Overlapcited by 105
  15. 2020Contrastive Learning with Adversarial Examplescited by 91
  16. 2019A Mutual Information Maximization Perspective of Language Representation Learningcited by 85
  17. 2021When Does Contrastive Learning Preserve Adversarial Robustness from Pretraining to Finetuning?cited by 76
  18. 2022Self-Ensemble Adversarial Training for Improved Robustnesscited by 38
  19. 2021Residual Relaxation for Multi-view Representation Learningcited by 30
  20. 2022Decoupled Adversarial Contrastive Learning for Self-supervised Adversarial Robustnesscited by 20
  21. 2019Understanding Adversarial Robustness Through Loss Landscape Geometriescited by 12
  22. 2018Interpreting Adversarial Robustness: A View from Decision Surface in Input Spacecited by 10

Built on by

Cited by year

Citing papers per year: 2023: 6; 2024: 10; 2025: 4.

20 papers.

Sort:

By area:

  1. 2024Robust CLIP: Unsupervised Adversarial Fine-Tuning of Vision Embeddings for Robust Large Vision-Language Modelscited by 119
  2. 2024One Prompt Word is Enough to Boost Adversarial Robustness for Pre-trained Vision-Language Modelscited by 48
  3. 2024Do Generated Data Always Help Contrastive Learning?cited by 25
  4. 2023Balance, Imbalance, and Rebalance: Understanding Robust Overfitting from a Minimax Game Perspectivecited by 18
  5. 2024Non-negative Contrastive Learningcited by 18
  6. 2023Adversarial Examples Are Not Real Featurescited by 16
  7. 2023Detection and Defense of Unlearnable Examplescited by 13
  8. 2023Enhancing Adversarial Contrastive Learning via Adversarial Invariant Regularizationcited by 10
  9. 2023Similarity Min-Max: Zero-Shot Day-Night Domain Adaptationcited by 10
  10. 2024GenView: Enhancing View Quality with Pretrained Generative Model for Self-Supervised Learningcited by 8
  11. 2024AdvQDet: Detecting Query-Based Adversarial Attacks with Adversarial Contrastive Prompt Tuningcited by 7
  12. 2024Benchmarking Robust Self-Supervised Learning Across Diverse Downstream Taskscited by 5
8 more
  1. 2025Graph Your Own Promptcited by 2
  2. 2024Exploiting Supervised Poison Vulnerability to Strengthen Self-Supervised Defensecited by 1
  3. 2024Revisiting Semi-supervised Adversarial Robustness via Noise-aware Online Robust Distillationcited by 1
  4. 2025Adversarial Robustness of Discriminative Self-Supervised Learning in Visioncited by 1
  5. 2025How does Labeling Error Impact Contrastive Learning? A Perspective from Data Dimensionality Reductioncited by 1
  6. 2023Generalization Bounds for Robust Contrastive Learning: From Theory to Practicecited by 0
  7. 2024ProFeAT: Projected Feature Adversarial Training for Self-Supervised Learning of Robust Representationscited by 0
  8. 2025Hierarchical Self-Supervised Adversarial Training for Robust Vision Models in Histopathologycited by 0