mlsanityData to 14 Oct 2026

Symmetric Cross Entropy for Robust Learning with Noisy Labels

arXiv 1908.06112 ↗First posted 16 Aug 2019cs.LG cs.CV stat.MLcited by 500

Not on the map. Why?

Abstract (arXiv)

Training accurate deep neural networks (DNNs) in the presence of noisy labels is an important and challenging task. Though a number of approaches have been proposed for learning with noisy labels, many open issues remain. In this paper, we show that DNN learning with Cross Entropy (CE) exhibits overfitting to noisy labels on some classes ("easy" classes), but more surprisingly, it also suffers from significant under learning on some other classes ("hard" classes). Intuitively, CE requires an extra term to facilitate learning of hard classes, and more importantly, this term should be noise tolerant, so as to avoid overfitting to noisy labels. Inspired by the symmetric KL-divergence, we propose the approach of Symmetric cross entropy Learning (SL), boosting CE symmetrically with a noise robust counterpart Reverse Cross Entropy (RCE). Our proposed SL approach simultaneously addresses both the under learning and overfitting problem of CE in the presence of noisy labels. We provide a theoretical analysis of SL and also empirically show, on a range of benchmark and real-world datasets, that SL outperforms state-of-the-art methods. We also show that SL can be easily incorporated into existing methods in order to further enhance their performance.

Key facts. Cited by 500 papers in the map (data through 14 Oct 2026). First posted on arXiv 16 Aug 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 98; 2022: 96; 2023: 90; 2024: 85; 2025: 81; 2026: 50.

500 papers.

Sort:

By area:

  1. 2023A Comprehensive Survey on Test-Time Adaptation under Distribution Shiftscited by 359
  2. 2021Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentationcited by 268
  3. 2021Learning with Noisy Labels Revisited: A Study Using Real-World Human Annotationscited by 256
  4. 2023Large Language Model as Attributed Training Data Generator: A Tale of Diversity and Biascited by 232
  5. 2022Backdoor Defense via Decoupling the Training Processcited by 153
  6. 2021Understanding and Improving Early Stopping for Learning with Noisy Labelscited by 113
  7. 2022Robust Mean Teacher for Continual and Gradual Test-Time Adaptationcited by 104
  8. 2021Learning with Feature-Dependent Label Noise: A Progressive Approachcited by 102
  9. 2022Robust Training under Label Noise by Over-parameterizationcited by 93
  10. 2021Dive into Ambiguity: Latent Distribution Mining and Pairwise Uncertainty Estimation for Facial Expression Recognitioncited by 89
  11. 2021FINE Samples for Learning with Noisy Labelscited by 82
  12. 2021Robust Audio-Visual Instance Discriminationcited by 80
38 more
  1. 2021A Simple Baseline for Semi-supervised Semantic Segmentation with Strong Data Augmentationcited by 79
  2. 2022Are All Losses Created Equal: A Neural Collapse Perspectivecited by 78
  3. 2021Generalized Jensen-Shannon Divergence Loss for Learning with Noisy Labelscited by 74
  4. 2021To Smooth or Not? When Label Smoothing Meets Noisy Labelscited by 70
  5. 2022ProGen: Progressive Zero-shot Dataset Generation via In-context Feedbackcited by 67
  6. 2021Clusterability as an Alternative to Anchor Points When Learning with Noisy Labelscited by 65
  7. 2023Guiding Pseudo-labels with Uncertainty Estimation for Source-free Unsupervised Domain Adaptationcited by 65
  8. 2023Universal Test-time Adaptation through Weight Ensembling, Diversity Weighting, and Prior Correctioncited by 65
  9. 2022PolyLoss: A Polynomial Expansion Perspective of Classification Loss Functionscited by 63
  10. 2021Learning Noise Transition Matrix from Only Noisy Labels via Total Variation Regularizationcited by 62
  11. 2022The Devil is in the Labels: Noisy Label Correction for Robust Scene Graph Generationcited by 60
  12. 2021Open-set Label Noise Can Improve Robustness Against Inherent Label Noisecited by 59
  13. 2022Part-based Pseudo Label Refinement for Unsupervised Person Re-identificationcited by 59
  14. 2021NGC: A Unified Framework for Learning with Open-World Noisy Datacited by 58
  15. 2021Detecting Corrupted Labels Without Training a Model to Predictcited by 58
  16. 2022Robust Cross-Modal Representation Learning with Progressive Self-Distillationcited by 55
  17. 2021Co-learning: Learning from Noisy Labels with Self-supervisioncited by 51
  18. 2022Self-Guided Noise-Free Data Generation for Efficient Zero-Shot Learningcited by 50
  19. 2022Instance-Dependent Label-Noise Learning with Manifold-Regularized Transition Matrix Estimationcited by 49
  20. 2021From Synthetic to Real: Unsupervised Domain Adaptation for Animal Pose Estimationcited by 48
  21. 2021Hard Sample Aware Noise Robust Learning for Histopathology Image Classificationcited by 48
  22. 2022ProMix: Combating Label Noise via Maximizing Clean Sample Utilitycited by 48
  23. 2023When Source-Free Domain Adaptation Meets Learning with Noisy Labelscited by 47
  24. 2021Asymmetric Loss Functions for Learning with Noisy Labelscited by 46
  25. 2021Improved Regularization and Robustness for Fine-tuning in Neural Networkscited by 45
  26. 2022Robust Contrastive Learning against Noisy Viewscited by 45
  27. 2023In Search of Lost Online Test-time Adaptation: A Surveycited by 44
  28. 2024A Theoretical Understanding of Self-Correction through In-context Alignmentcited by 44
  29. 2023Backdoor Defense via Adaptively Splitting Poisoned Datasetcited by 43
  30. 2021LongReMix: Robust Learning with High Confidence Samples in a Noisy Label Environmentcited by 40
  31. 2022Mitigating Memorization of Noisy Labels by Clipping the Model Predictioncited by 40
  32. 2021Learning with Noisy Labels via Sparse Regularizationcited by 39
  33. 2022BMD: A General Class-balanced Multicentric Dynamic Prototype Strategy for Source-free Domain Adaptationcited by 39
  34. 2021Joint Negative and Positive Learning for Noisy Labelscited by 34
  35. 2021Learning from Noisy Labels for Entity-Centric Information Extractioncited by 34
  36. 2021Analysis and Applications of Class-wise Robustness in Adversarial Trainingcited by 33
  37. 2022Self-Filtering: A Noise-Aware Sample Selection for Label Noise with Confidence Penalizationcited by 33
  38. 2023FedNoRo: Towards Noise-Robust Federated Learning by Addressing Class Imbalance and Label Noise Heterogeneitycited by 33

50 most cited of 500 All 500 (JSON)