mlsanityData to 7 Oct 2026

MixMatch: A Holistic Approach to Semi-Supervised Learning

arXiv 1905.02249 ↗First posted 6 May 2019cs.LG cs.AI cs.CVcited by 1,390

Not on the map. Why?

Abstract (arXiv)

Semi-supervised learning has proven to be a powerful paradigm for leveraging unlabeled data to mitigate the reliance on large labeled datasets. In this work, we unify the current dominant approaches for semi-supervised learning to produce a new algorithm, MixMatch, that works by guessing low-entropy labels for data-augmented unlabeled examples and mixing labeled and unlabeled data using MixUp. We show that MixMatch obtains state-of-the-art results by a large margin across many datasets and labeled data amounts. For example, on CIFAR-10 with 250 labels, we reduce error rate by a factor of 4 (from 38% to 11%) and by a factor of 2 on STL-10. We also demonstrate how MixMatch can help achieve a dramatically better accuracy-privacy trade-off for differential privacy. Finally, we perform an ablation study to tease apart which components of MixMatch are most important for its success.

Key facts. Cited by 1,390 papers in the map (data through 7 Oct 2026). First posted on arXiv 6 May 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2020: 1; 2021: 325; 2022: 337; 2023: 316; 2024: 227; 2025: 123; 2026: 61.

1,390 papers.

Sort:

By area:

  1. 2021FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labelingcited by 537
  2. 2021Semi-Supervised Semantic Segmentation with Cross Pseudo Supervisioncited by 462
  3. 2023Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervisioncited by 445
  4. 2023A Comprehensive Survey on Test-Time Adaptation under Distribution Shiftscited by 353
  5. 2021End-to-End Semi-Supervised Object Detection with Soft Teachercited by 315
  6. 2021Unbiased Teacher for Semi-Supervised Object Detectioncited by 305
  7. 2021In Defense of Pseudo-Labeling: An Uncertainty-Aware Pseudo-label Selection Framework for Semi-Supervised Learningcited by 301
  8. 2021DatasetGAN: Efficient Labeled Data Factory with Minimal Human Effortcited by 264
  9. 2021Learning with Noisy Labels Revisited: A Study Using Real-World Human Annotationscited by 253
  10. 2021A Survey on Deep Semi-supervised Learningcited by 250
  11. 2023A Cookbook of Self-Supervised Learningcited by 248
  12. 2022FreeMatch: Self-adaptive Thresholding for Semi-supervised Learningcited by 232
38 more
  1. 2021One Million Scenes for Autonomous Driving: ONCE Datasetcited by 218
  2. 2021GPT3Mix: Leveraging Large-scale Language Models for Text Augmentationcited by 185
  3. 2021ST++: Make Self-training Work Better for Semi-supervised Semantic Segmentationcited by 184
  4. 2022Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentationcited by 177
  5. 2021CReST: A Class-Rebalancing Self-Training Framework for Imbalanced Semi-Supervised Learningcited by 174
  6. 2021The Values Encoded in Machine Learning Researchcited by 172
  7. 2021Open-World Semi-Supervised Learningcited by 165
  8. 2025The Unreasonable Effectiveness of Entropy Minimization in LLM Reasoningcited by 160
  9. 2021Self-supervised Augmentation Consistency for Adapting Semantic Segmentationcited by 149
  10. 2022Backdoor Defense via Decoupling the Training Processcited by 149
  11. 2021Dash: Semi-Supervised Learning with Dynamic Thresholdingcited by 147
  12. 2021Semantic Segmentation with Generative Models: Semi-Supervised Learning and Strong Out-of-Domain Generalizationcited by 141
  13. 2023SoftMatch: Addressing the Quantity-Quality Trade-off in Semi-supervised Learningcited by 133
  14. 2021Semi-Supervised Semantic Segmentation with Pixel-Level Contrastive Learning from a Class-wise Memory Bankcited by 125
  15. 2021Cycle Self-Training for Domain Adaptationcited by 120
  16. 2021Neighborhood Contrastive Learning for Novel Class Discoverycited by 119
  17. 2022SimMatch: Semi-supervised Learning with Similarity Matchingcited by 118
  18. 2021Image-to-Image Translation: Methods and Applicationscited by 116
  19. 2021Data Collection and Quality Challenges in Deep Learning: A Data-Centric AI Perspectivecited by 116
  20. 2021Humble Teachers Teach Better Students for Semi-Supervised Object Detectioncited by 115
  21. 2021Semi-Supervised Learning of Visual Features by Non-Parametrically Predicting View Assignments with Support Samplescited by 114
  22. 2021Perturbed and Strict Mean Teachers for Semi-supervised Semantic Segmentationcited by 113
  23. 2022Parametric Classification for Generalized Category Discovery: A Baseline Studycited by 113
  24. 2021Instant-Teaching: An End-to-End Semi-Supervised Object Detection Frameworkcited by 112
  25. 2021Semi-supervised Semantic Segmentation with Directional Context-aware Consistencycited by 112
  26. 2021Understanding and Improving Early Stopping for Learning with Noisy Labelscited by 112
  27. 2024Entropy is not Enough for Test-Time Adaptation: From the Perspective of Disentangled Factorscited by 106
  28. 2021SPICE: Semantic Pseudo-labeling for Image Clusteringcited by 104
  29. 2021Recent advances and clinical applications of deep learning in medical image analysiscited by 104
  30. 2021AdaMatch: A Unified Approach to Semi-Supervised Learning and Domain Adaptationcited by 104
  31. 2021Novel Visual Category Discovery with Dual Ranking Statistics and Mutual Knowledge Distillationcited by 104
  32. 2023A Comprehensive Survey on Source-free Domain Adaptationcited by 104
  33. 2021An Empirical Survey of Data Augmentation for Limited Data Learning in NLPcited by 101
  34. 2021Semi-Supervised Semantic Segmentation via Adaptive Equalization Learningcited by 101
  35. 2021Contrastive Learning with Stronger Augmentationscited by 100
  36. 2025Maximizing Confidence Alone Improves Reasoningcited by 97
  37. 2022SURF: Semi-supervised Reward Learning with Data Augmentation for Feedback-efficient Preference-based Reinforcement Learningcited by 96
  38. 2022USB: A Unified Semi-supervised Learning Benchmark for Classificationcited by 95

50 most cited of 1,390 All 1,390 (JSON)