mlsanityData to 7 Oct 2026

Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss

arXiv 1906.07413 ↗First posted 18 Jun 2019cs.LG cs.CV stat.MLcited by 987

Not on the map. Why?

Cited by 29 in Diffusion models, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Deep learning algorithms can fare poorly when the training dataset suffers from heavy class-imbalance but the testing criterion requires good generalization on less frequent classes. We design two novel methods to improve performance in such scenarios. First, we propose a theoretically-principled label-distribution-aware margin (LDAM) loss motivated by minimizing a margin-based generalization bound. This loss replaces the standard cross-entropy objective during training and can be applied with prior strategies for training with class-imbalance such as re-weighting or re-sampling. Second, we propose a simple, yet effective, training schedule that defers re-weighting until after the initial stage, allowing the model to learn an initial representation while avoiding some of the complications associated with re-weighting or re-sampling. We test our methods on several benchmark vision tasks including the real-world imbalanced dataset iNaturalist 2018. Our experiments show that either of these methods alone can already improve over existing techniques and their combination achieves even better performance gains.

Key facts. Cited by 987 papers in the map (data through 7 Oct 2026). First posted on arXiv 18 Jun 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 143; 2022: 178; 2023: 213; 2024: 186; 2025: 138; 2026: 129.

987 papers.

Sort:

By area:

  1. 2021Just Train Twice: Improving Group Robustness without Training Group Informationcited by 522
  2. 2021Deep Long-Tailed Learning: A Surveycited by 416
  3. 2022Last Layer Re-Training is Sufficient for Robustness to Spurious Correlationscited by 379
  4. 2022Language Models are Realistic Tabular Data Generatorscited by 290
  5. 2021IconQA: A New Benchmark for Abstract Diagram Understanding and Visual Language Reasoningcited by 262
  6. 2021Parametric Contrastive Learningcited by 217
  7. 2021Improving Calibration for Long-Tailed Recognitioncited by 212
  8. 2022Improving Out-of-Distribution Robustness via Selective Augmentationcited by 202
  9. 2021Distribution Alignment: A Unified Framework for Long-tail Visual Recognitioncited by 198
  10. 2021BENDR: using transformers and a contrastive self-supervised learning task to learn from massive amounts of EEG datacited by 194
  11. 2021CReST: A Class-Rebalancing Self-Training Framework for Imbalanced Semi-Supervised Learningcited by 174
  12. 2021Exploring Deep Neural Networks via Layer-Peeled Model: Minority Collapse in Imbalanced Trainingcited by 166
38 more
  1. 2021Open-World Semi-Supervised Learningcited by 165
  2. 2021Delving into Deep Imbalanced Regressioncited by 157
  3. 2022Balanced Contrastive Learning for Long-Tailed Visual Recognitioncited by 143
  4. 2021Bipartite Graph Network with Adaptive Message Passing for Unbiased Scene Graph Generationcited by 137
  5. 2021Targeted Supervised Contrastive Learning for Long-Tailed Recognitioncited by 132
  6. 2021Contrastive Learning based Hybrid Networks for Long-Tailed Image Classificationcited by 130
  7. 2021Self-supervised Learning is More Robust to Dataset Imbalancecited by 122
  8. 2022Diffusion Probabilistic Model Made Slimcited by 121
  9. 2023Change is Hard: A Closer Look at Subpopulation Shiftcited by 119
  10. 2022Federated Learning with Label Distribution Skew via Logits Calibrationcited by 108
  11. 2022Long-Tailed Recognition via Weight Balancingcited by 107
  12. 2022Retrieval Augmented Classification for Long-Tail Visual Recognitioncited by 106
  13. 2022Animal Kingdom: A Large and Diverse Dataset for Animal Behavior Understandingcited by 102
  14. 2021Semi-Supervised Semantic Segmentation via Adaptive Equalization Learningcited by 101
  15. 2021The Majority Can Help The Minority: Context-rich Minority Oversampling for Long-tailed Classificationcited by 100
  16. 2021DENSE: Data-Free One-Shot Federated Learningcited by 100
  17. 2022Inducing Neural Collapse in Imbalanced Learning: Do We Really Need a Learnable Classifier at the End of Deep Neural Network?cited by 100
  18. 2021MetaSAug: Meta Semantic Augmentation for Long-Tailed Visual Recognitioncited by 99
  19. 2021ResLT: Residual Learning for Long-tailed Recognitioncited by 96
  20. 2021Self-Supervised Aggregation of Diverse Experts for Test-Agnostic Long-Tailed Recognitioncited by 95
  21. 2021Influence-Balanced Loss for Imbalanced Visual Classificationcited by 91
  22. 2022Balanced MSE for Imbalanced Visual Regressioncited by 89
  23. 2021ACE: Ally Complementary Experts for Solving Long-Tailed Recognition in One-Shotcited by 86
  24. 2022Debiased Learning from Naturally Imbalanced Pseudo-Labelscited by 85
  25. 2021Distributional Robustness Loss for Long-tail Learningcited by 84
  26. 2022Imbalance Trouble: Revisiting Neural-Collapse Geometrycited by 80
  27. 2021Detecting Human-Object Interaction via Fabricated Compositional Learningcited by 79
  28. 2021FASA: Feature Augmentation and Sampling Adaptation for Long-Tailed Instance Segmentationcited by 76
  29. 2022A Survey on Long-Tailed Visual Recognitioncited by 75
  30. 2023Long-tailed Visual Recognition via Gaussian Clouded Logit Adjustmentcited by 75
  31. 2021Label-Imbalanced and Group-Sensitive Classification under Overparameterizationcited by 74
  32. 2021Self Supervision to Distillation for Long-Tailed Visual Recognitioncited by 73
  33. 2021Distilling Virtual Examples for Long-tailed Recognitioncited by 72
  34. 2024Preserving Fairness Generalization in Deepfake Detectioncited by 70
  35. 2021Learning Fast Sample Re-weighting Without Reward Datacited by 69
  36. 2022LPT: Long-tailed Prompt Tuning for Image Classificationcited by 69
  37. 2022Tackling Data Heterogeneity in Federated Learning with Class Prototypescited by 66
  38. 2021Learning of Visual Relations: The Devil is in the Tailscited by 65

50 most cited of 987 All 987 (JSON)