mlsanityData to 7 Oct 2026

Maximum-Entropy Adversarial Data Augmentation for Improved Generalization and Robustness

arXiv 2010.08001 ↗First posted 15 Oct 2020cs.LG cs.CVcited by 121

Not on the map. Why?

Abstract (arXiv)

Adversarial data augmentation has shown promise for training robust deep neural networks against unforeseen data shifts or corruptions. However, it is difficult to define heuristics to generate effective fictitious target distributions containing "hard" adversarial perturbations that are largely different from the source distribution. In this paper, we propose a novel and effective regularization term for adversarial data augmentation. We theoretically derive it from the information bottleneck principle, which results in a maximum-entropy formulation. Intuitively, this regularization term encourages perturbing the underlying source distribution to enlarge predictive uncertainty of the current model, so that the generated "hard" adversarial perturbations can improve the model robustness during training. Experimental results on three standard benchmarks demonstrate that our method consistently outperforms the existing state of the art by a statistically significant margin.

Key facts. Cited by 121 papers in the map (data through 7 Oct 2026). First posted on arXiv 15 Oct 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 14; 2022: 21; 2023: 37; 2024: 26; 2025: 15; 2026: 8.

121 papers.

Sort:

By area:

  1. 2023SmoothLLM: Defending Large Language Models Against Jailbreaking Attackscited by 434
  2. 2022Improving Out-of-Distribution Robustness via Selective Augmentationcited by 202
  3. 2021Learning to Diversify for Single Domain Generalizationcited by 170
  4. 2021Progressive Domain Expansion Network for Single Domain Generalizationcited by 96
  5. 2021AugMax: Adversarial Composition of Random Augmentations for Robust Trainingcited by 89
  6. 2021Adversarially Adaptive Normalization for Single Domain Generalizationcited by 85
  7. 2022Adversarial Style Augmentation for Domain Generalized Urban-Scene Segmentationcited by 69
  8. 2022DrugOOD: Out-of-Distribution (OOD) Dataset Curator and Benchmark for AI-aided Drug Discovery -- A Focus on Affinity Prediction Problems with Noise Annotationscited by 54
  9. 2021Non-Transferable Learning: A New Approach for Model Ownership Verification and Applicability Authorizationcited by 50
  10. 2023Training on Thin Air: Improve Image Classification with Generated Datacited by 50
  11. 2021A Systematic Review of Robustness in Deep Learning for Computer Vision: Mind the gap?cited by 48
  12. 2023Meta-causal Learning for Single Domain Generalizationcited by 43
38 more
  1. 2021Uncertainty-guided Model Generalization to Unseen Domainscited by 41
  2. 2023Robust and Interpretable Medical Image Classifiers via Concept Bottleneck Modelscited by 40
  3. 2022Attention Consistency on Visual Corruptions for Single-Source Domain Generalizationcited by 37
  4. 2023StyLIP: Multi-Scale Style-Conditioned Prompt Learning for CLIP-based Domain Generalizationcited by 36
  5. 2023Progressive Random Convolutions for Single Domain Generalizationcited by 33
  6. 2024Improving Single Domain-Generalized Object Detection: A Focus on Diversification and Alignmentcited by 29
  7. 2021Ranking Distance Calibration for Cross-Domain Few-Shot Learningcited by 27
  8. 2024Prompt-Driven Dynamic Object-Centric Learning for Single Domain Generalizationcited by 24
  9. 2021Taxonomy of Machine Learning Safety: A Survey and Primercited by 22
  10. 2022Style-Hallucinated Dual Consistency Learning: A Unified Framework for Visual Domain Generalizationcited by 22
  11. 2023A Survey on the Robustness of Computer Vision Models against Common Corruptionscited by 20
  12. 2024Distribution-Aware Data Expansion with Diffusion Modelscited by 20
  13. 2023Domain Generalization via Balancing Training Difficulty and Model Capabilitycited by 19
  14. 2021Out-of-Domain Generalization from a Single Source: An Uncertainty Quantification Approachcited by 18
  15. 2024Robust Domain Generalization for Multi-modal Object Recognitioncited by 18
  16. 2022Not Just Pretty Pictures: Toward Interventional Data Augmentation Using Text-to-Image Generatorscited by 17
  17. 2022Bootstrap Generalization Ability from Loss Landscape Perspectivecited by 16
  18. 2023Adversarial AutoMixupcited by 16
  19. 2023AdvST: Revisiting Data Augmentations for Single Domain Generalizationcited by 16
  20. 2022Gate-Shift-Fuse for Video Action Recognitioncited by 15
  21. 2022Adversarial Unlearning: Reducing Confidence Along Adversarial Directionscited by 14
  22. 2023Towards Artificial General Intelligence (AGI) in the Internet of Things (IoT): Opportunities and Challengescited by 14
  23. 2022Learning Instance-Specific Adaptation for Cross-Domain Segmentationcited by 13
  24. 2022Learning Robust Representations for Continual Relation Extraction via Adversarial Class Augmentationcited by 13
  25. 2022GSMFlow: Generation Shifts Mitigating Flow for Generalized Zero-Shot Learningcited by 12
  26. 2023Fourier Test-time Adaptation with Multi-level Consistency for Robust Classificationcited by 12
  27. 2025Robustness in Large Language Models: A Survey of Mitigation Strategies and Evaluation Metricscited by 12
  28. 2022Global-Local Regularization Via Distributional Robustnesscited by 11
  29. 2024PhysAug: A Physical-guided and Frequency-based Data Augmentation for Single-Domain Generalized Object Detectioncited by 11
  30. 2021Cooperative Training and Latent Space Data Augmentation for Robust Medical Image Segmentationcited by 9
  31. 2023Complementary Domain Adaptation and Generalization for Unsupervised Continual Domain Shift Learningcited by 9
  32. 2023Activate and Reject: Towards Safe Domain Generalization under Category Shiftcited by 9
  33. 2023Visual Data-Type Understanding does not emerge from Scaling Vision-Language Modelscited by 9
  34. 2022GeoECG: Data Augmentation via Wasserstein Geodesic Perturbation for Robust Electrocardiogram Predictioncited by 8
  35. 2022A.I. Robustness: a Human-Centered Perspective on Technological Challenges and Opportunitiescited by 8
  36. 2022Adversarial Purification with the Manifold Hypothesiscited by 8
  37. 2023Learning Class and Domain Augmentations for Single-Source Open-Domain Generalizationcited by 8
  38. 2024Generalizable Person Re-identification via Balancing Alignment and Uniformitycited by 8

50 most cited of 121 All 121 (JSON)