mlsanityData to 7 Oct 2026

Fairness Improves Learning from Noisily Labeled Long-Tailed Data

arXiv 2303.12291 ↗First posted 22 Mar 2023cs.LGcited by 5

Not on the map. Why?

Abstract (arXiv)

Both long-tailed and noisily labeled data frequently appear in real-world applications and impose significant challenges for learning. Most prior works treat either problem in an isolated way and do not explicitly consider the coupling effects of the two. Our empirical observation reveals that such solutions fail to consistently improve the learning when the dataset is long-tailed with label noise. Moreover, with the presence of label noise, existing methods do not observe universal improvements across different sub-populations; in other words, some sub-populations enjoyed the benefits of improved accuracy at the cost of hurting others. Based on these observations, we introduce the Fairness Regularizer (FR), inspired by regularizing the performance gap between any two sub-populations. We show that the introduced fairness regularizer improves the performances of sub-populations on the tail and the overall learning performance. Extensive experiments demonstrate the effectiveness of the proposed solution when complemented with certain existing popular robust or class-balanced methods.

Key facts. Cited by 5 papers in the map (data through 7 Oct 2026). Builds on 46 of its 74 references that are in the map. First posted on arXiv 22 Mar 2023.

Builds on

46 of 74 references.

References · 46

  1. 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
  2. 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,040
  3. 2017Axiomatic Attribution for Deep Networkscited by 3,350
  4. 2016Equality of Opportunity in Supervised Learningcited by 2,000
  5. 2017Accurate, Large Minibatch SGD: Training ImageNet in 1 Hourcited by 1,479
  6. 2019Class-Balanced Loss Based on Effective Number of Samplescited by 1,239
  7. 2018Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labelscited by 1,053
  8. 2016Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approachcited by 663
  9. 2020Long-tail learning via logit adjustmentcited by 605
  10. 2020What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimationcited by 492
  11. 2021Deep Long-Tailed Learning: A Surveycited by 416
  12. 2020Early-Learning Regularization Prevents Memorization of Noisy Labelscited by 414
34 more
  1. 2020Balanced Meta-Softmax for Long-Tailed Visual Recognitioncited by 397
  2. 2020Combating noisy labels by agreement: A joint training method with co-regularizationcited by 289
  3. 2019Towards Efficient Data Valuation Based on the Shapley Valuecited by 279
  4. 2021Learning with Noisy Labels Revisited: A Study Using Real-World Human Annotationscited by 253
  5. 2022Mitigating Neural Network Overconfidence with Logit Normalizationcited by 251
  6. 2020Does label smoothing mitigate label noise?cited by 204
  7. 2017CleanNet: Transfer Learning for Scalable Image Classifier Training with Label Noisecited by 181
  8. 2020Learning with Instance-Dependent Label Noise: A Sample Sieve Approachcited by 158
  9. 2019NLNL: Negative Learning for Noisy Labelscited by 151
  10. 2019Inherent Tradeoffs in Learning Fair Representationscited by 144
  11. 2019Peer Loss Functions: Learning from Noisy Labels without Knowing Noise Ratescited by 142
  12. 2015Learning with Symmetric Label Noise: The Importance of Being Unhingedcited by 120
  13. 2022Robust Training under Label Noise by Over-parameterizationcited by 93
  14. 2021Influence-Balanced Loss for Imbalanced Visual Classificationcited by 91
  15. 2020A Second-Order Approach to Learning with Instance-Dependent Label Noisecited by 90
  16. 2021Sample Selection with Uncertainty of Losses for Learning with Noisy Labelscited by 79
  17. 2013Classification with Asymmetric Label Noise: Consistency and Maximal Denoisingcited by 67
  18. 2020Fair Classification with Group-Dependent Label Noisecited by 67
  19. 2021Clusterability as an Alternative to Anchor Points When Learning with Noisy Labelscited by 65
  20. 2021Adaptive Early-Learning Correction for Segmentation from Noisy Annotationscited by 61
  21. 2021Open-set Label Noise Can Improve Robustness Against Inherent Label Noisecited by 58
  22. 2021Detecting Corrupted Labels Without Training a Model to Predictcited by 57
  23. 2020When Optimizing f-divergence is Robust with Label Noisecited by 44
  24. 2021Robust Long-Tailed Learning under Label Noisecited by 35
  25. 2021The Rich Get Richer: Disparate Impact of Semi-Supervised Learningcited by 28
  26. 2022Beyond Images: Label Noise Transition Matrix Estimation for Tasks with Lower-Quality Featurescited by 27
  27. 2022Understanding Instance-Level Impact of Fairness Constraintscited by 27
  28. 2021Mitigating Memorization of Noisy Labels via Regularization between Representationscited by 26
  29. 2021Understanding Instance-Level Label Noise: Disparate Impacts and Treatmentscited by 22
  30. 2021Learning From Long-Tailed Data With Noisy Labelscited by 17
  31. 2019Fairness With Minimal Harm: A Pareto-Optimal Approach For Healthcarecited by 16
  32. 2022PADDLES: Phase-Amplitude Spectrum Disentangled Early Stopping for Learning with Noisy Labelscited by 10
  33. 2021NoiLIn: Improving Adversarial Training and Correcting Stereotype of Noisy Labelscited by 8
  34. 2021Instance Correction for Learning with Open-set Noisy Labelscited by 7

Built on by

Cited by year

Citing papers per year: 2023: 2; 2024: 0; 2025: 3.

5 papers.

  1. 2023Unprocessing Seven Years of Algorithmic Fairnesscited by 18
  2. 2025When VLMs Meet Image Classification: Test Sets Renovation via Missing Label Identificationcited by 5
  3. 2025SelectMix: Enhancing Label Noise Robustness through Targeted Sample Mixingcited by 5
  4. 2023Mitigating Source Bias for Fairer Weak Supervisioncited by 3
  5. 2025Dual-granularity Sinkhorn Distillation for Enhanced Learning from Long-tailed Noisy Datacited by 1