mlsanityData to 7 Oct 2026

Asymmetric Loss Functions for Learning with Noisy Labels

arXiv 2106.03110 ↗First posted 6 Jun 2021cs.LG cs.CVcited by 46

Not on the map. Why?

Abstract (arXiv)

Robust loss functions are essential for training deep neural networks with better generalization power in the presence of noisy labels. Symmetric loss functions are confirmed to be robust to label noise. However, the symmetric condition is overly restrictive. In this work, we propose a new class of loss functions, namely textitasymmetric loss functions, which are robust to learning with noisy labels for various types of noise. We investigate general theoretical properties of asymmetric loss functions, including classification calibration, excess risk bound, and noise tolerance. Meanwhile, we introduce the asymmetry ratio to measure the asymmetry of a loss function. The empirical results show that a higher ratio would provide better noise tolerance. Moreover, we modify several commonly-used loss functions and establish the necessary and sufficient conditions for them to be asymmetric. Experimental results on benchmark datasets demonstrate that asymmetric loss functions can outperform state-of-the-art methods. The code is available at hrefhttps://github.com/hitcszx/ALFshttps://github.com/hitcszx/ALFs

Key facts. Cited by 46 papers in the map (data through 7 Oct 2026). Builds on 19 of its 39 references that are in the map. First posted on arXiv 6 Jun 2021.

Builds on

19 of 39 references.

References · 19

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2016Understanding deep learning requires rethinking generalizationcited by 1,657
  3. 2018Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labelscited by 994
  4. 2017A Closer Look at Memorization in Deep Networkscited by 936
  5. 2016Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approachcited by 663
  6. 2017MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labelscited by 652
  7. 2020Learning from Noisy Labels with Deep Neural Networks: A Surveycited by 580
  8. 2019Symmetric Cross Entropy for Robust Learning with Noisy Labelscited by 499
  9. 2017Robust Loss Functions under Label Noise for Deep Neural Networkscited by 441
  10. 2017WebVision Database: Visual Learning and Understanding from Web Datacited by 262
  11. 2020Normalized Loss Functions for Deep Learning with Noisy Labelscited by 257
  12. 2019NLNL: Negative Learning for Noisy Labelscited by 151
7 more
  1. 2019Peer Loss Functions: Learning from Noisy Labels without Knowing Noise Ratescited by 142
  2. 2015Learning with Symmetric Label Noise: The Importance of Being Unhingedcited by 120
  3. 2019An Improved Analysis of Training Over-parameterized Deep Neural Networkscited by 120
  4. 2019Curriculum Loss: Robust Learning and Generalization against Label Corruptioncited by 87
  5. 2017Learning with Bounded Instance- and Label-dependent Label Noisecited by 57
  6. 2014Making Risk Minimization Tolerant to Label Noisecited by 54
  7. 2019On Symmetric Losses for Learning from Corrupted Labelscited by 53

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 3; 2022: 6; 2023: 10; 2024: 9; 2025: 12; 2026: 6.

46 papers.

Sort:

By area:

  1. 2022Mitigating Memorization of Noisy Labels by Clipping the Model Predictioncited by 40
  2. 2021Learning with Noisy Labels via Sparse Regularizationcited by 38
  3. 2023SILT: Shadow-aware Iterative Label Tuning for Learning to Detect Shadows from Noisy Labelscited by 19
  4. 2024CLIPCleaner: Cleaning Noisy Labels with CLIPcited by 16
  5. 2022Scalable Penalized Regression for Noise Detection in Learning with Noisy Labelscited by 15
  6. 2021Label Distributionally Robust Losses for Multi-class Classification: Consistency, Robustness and Adaptivitycited by 11
  7. 2023Mitigating Label Noise through Data Ambiguationcited by 10
  8. 2022Instance-dependent Label Distribution Estimation for Learning with Label Noisecited by 9
  9. 2023Latent Class-Conditional Noise Modelcited by 9
  10. 2023Manifold DivideMix: A Semi-Supervised Contrastive Learning Framework for Severe Label Noisecited by 9
  11. 2021Sample Prior Guided Robust Model Learning to Suppress Noisy Labelscited by 8
  12. 2023Dynamics-Aware Loss for Learning with Label Noisecited by 8
34 more
  1. 2023Fine tuning Pre trained Models for Robustness Under Noisy Labelscited by 8
  2. 2025ε-Softmax: Approximating One-Hot Vectors for Mitigating Label Noisecited by 6
  3. 2024An Embedding is Worth a Thousand Noisy Labelscited by 5
  4. 2022A Gift from Label Smoothing: Robust Training with Adaptive Label Smoothing via Auxiliary Classifier under Label Noisecited by 4
  5. 2022Prototype-Anchored Learning for Learning with Imperfect Annotationscited by 4
  6. 2024Optimized Gradient Clipping for Noisy Label Learningcited by 4
  7. 2023Logistic-Normal Likelihoods for Heteroscedastic Label Noisecited by 3
  8. 2023Enhancing Noise-Robust Losses for Large-Scale Noisy Data Learningcited by 3
  9. 2024Active Negative Loss: A Robust Framework for Learning with Noisy Labelscited by 3
  10. 2025Enhanced Sample Selection with Confidence Tracking: Identifying Correctly Labeled yet Hard-to-Learn Samples in Noisy Datacited by 3
  11. 2025Robust Classification with Noisy Labels Based on Posterior Maximizationcited by 2
  12. 2024Unleashing the Potential of Open-set Noisy Samples Against Label Noise for Medical Image Classificationcited by 1
  13. 2024Noisy Early Stopping for Noisy Labelscited by 1
  14. 2024Implicit to Explicit Entropy Regularization: Benchmarking ViT Fine-tuning under Noisy Labelscited by 1
  15. 2025Think Twice before Adaptation: Improving Adaptability of DeepFake Detection via Online Test-Time Adaptationcited by 1
  16. 2025Drainage: A Unifying Framework for Addressing Class Uncertaintycited by 1
  17. 2026Symmetrization of Loss Functions for Robust Training of Neural Networks in the Presence of Noisy Labelscited by 1
  18. 2022AugLoss: A Robust Augmentation-based Fine Tuning Methodologycited by 0
  19. 2023A Curriculum View of Robust Loss Functionscited by 0
  20. 2023Neural Boundingcited by 0
  21. 2024Machine Learning in Proton Exchange Membrane Water Electrolysis -- Part I: A Knowledge-Integrated Frameworkcited by 0
  22. 2024Distribution-aware Noisy-label Crack Segmentationcited by 0
  23. 2025Set a Thief to Catch a Thief: Combating Label Noise through Noisy Meta Learningcited by 0
  24. 2025Bridging Weakly-Supervised Learning and VLM Distillation: Noisy Partial Label Learning for Efficient Downstream Adaptationcited by 0
  25. 2025Robust Classification under Noisy Labels: A Geometry-Aware Reliability Framework for Foundation Modelscited by 0
  26. 2025Learning to Forget with Information Divergence Reweighted Objectives for Noisy Labelscited by 0
  27. 2025Integrating SAM Supervision for 3D Weakly Supervised Point Cloud Segmentationcited by 0
  28. 2025Hard Samples, Bad Labels: Robust Loss Functions That Know When to Back Offcited by 0
  29. 2025Learning to Retrieve with Weakened Labels: Robust Training under Label Noisecited by 0
  30. 2026Minimax Generalized Cross-Entropycited by 0
  31. 2026Intrinsic Gradient Suppression for Label-Noise Prompt Tuning in Vision-Language Modelscited by 0
  32. 2026Conveyance: A Versatile Framework for Learning in Structured Class Spacescited by 0
  33. 2026XGRVFL-MV: Residual-Coupled Graph-Embedded Multi-View Random Vector Functional Link Network with FleXi Guardian Losscited by 0
  34. 2026Learning Sign Language Recognition under Label Noise: A Study of Noise-Robust Losses for Isolated and Continuous Settingscited by 0