mlsanityData to 14 Oct 2026

Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentation

arXiv 2208.09910 ↗First posted 21 Aug 2022cs.CVcited by 178

Not on the map. Why?

Abstract (arXiv)

In this work, we revisit the weak-to-strong consistency framework, popularized by FixMatch from semi-supervised classification, where the prediction of a weakly perturbed image serves as supervision for its strongly perturbed version. Intriguingly, we observe that such a simple pipeline already achieves competitive results against recent advanced works, when transferred to our segmentation scenario. Its success heavily relies on the manual design of strong data augmentations, however, which may be limited and inadequate to explore a broader perturbation space. Motivated by this, we propose an auxiliary feature perturbation stream as a supplement, leading to an expanded perturbation space. On the other, to sufficiently probe original image-level augmentations, we present a dual-stream perturbation technique, enabling two strong views to be simultaneously guided by a common weak view. Consequently, our overall Unified Dual-Stream Perturbations approach (UniMatch) surpasses all existing methods significantly across all evaluation protocols on the Pascal, Cityscapes, and COCO benchmarks. Its superiority is also demonstrated in remote sensing interpretation and medical image analysis. We hope our reproduced FixMatch and our results can inspire more future works. Code and logs are available at https://github.com/LiheYoung/UniMatch.

Key facts. Cited by 178 papers in the map (data through 14 Oct 2026). Builds on 64 of its 82 references that are in the map. First posted on arXiv 21 Aug 2022.

Builds on

64 of 82 references.

References · 64

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  3. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
  4. 2018Representation Learning with Contrastive Predictive Codingcited by 7,777
  5. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
  6. 2022Generative Adversarial Networkscited by 5,627
  7. 2014Fully Convolutional Networks for Semantic Segmentationcited by 4,859
  8. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,444
  9. 2018Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentationcited by 3,223
  10. 2019CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Featurescited by 2,884
  11. 2020Exploring Simple Siamese Representation Learningcited by 2,866
  12. 2020Unsupervised Learning of Visual Features by Contrasting Cluster Assignmentscited by 2,857
38 more
  1. 2016Pyramid Scene Parsing Networkcited by 2,479
  2. 2016Xception: Deep Learning with Depthwise Separable Convolutionscited by 2,332
  3. 2020FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidencecited by 2,085
  4. 2017Improved Regularization of Convolutional Neural Networks with Cutoutcited by 1,628
  5. 2019MixMatch: A Holistic Approach to Semi-Supervised Learningcited by 1,391
  6. 2019Self-training with Noisy Student improves ImageNet classificationcited by 1,311
  7. 2019Unsupervised Data Augmentation for Consistency Trainingcited by 1,060
  8. 2016Temporal Ensembling for Semi-Supervised Learningcited by 1,057
  9. 2017Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learningcited by 920
  10. 2017Deep Mutual Learningcited by 628
  11. 2018ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentationcited by 623
  12. 2021FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labelingcited by 540
  13. 2019Pseudo-Labeling and Confirmation Bias in Deep Semi-Supervised Learningcited by 510
  14. 2021Semi-Supervised Semantic Segmentation with Cross Pseudo Supervisioncited by 463
  15. 2016Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learningcited by 441
  16. 2019Uncertainty-aware Self-ensembling Model for Semi-supervised 3D Left Atrium Segmentationcited by 379
  17. 2020Semi-Supervised Semantic Segmentation with Cross-Consistency Trainingcited by 376
  18. 2020Meta Pseudo Labelscited by 374
  19. 2020Rethinking Pre-training and Self-trainingcited by 352
  20. 2019ReMixMatch: Semi-Supervised Learning with Distribution Alignment and Augmentation Anchoringcited by 349
  21. 2021End-to-End Semi-Supervised Object Detection with Soft Teachercited by 316
  22. 2021Unbiased Teacher for Semi-Supervised Object Detectioncited by 306
  23. 2018Context Encoding for Semantic Segmentationcited by 290
  24. 2022FreeMatch: Self-adaptive Thresholding for Semi-supervised Learningcited by 232
  25. 2022Semi-Supervised Semantic Segmentation Using Unreliable Pseudo-Labelscited by 218
  26. 2021ST++: Make Self-training Work Better for Semi-supervised Semantic Segmentationcited by 184
  27. 2020PseudoSeg: Designing Pseudo Labels for Semantic Segmentationcited by 180
  28. 2020CoMatch: Semi-supervised Learning with Contrastive Graph Regularizationcited by 167
  29. 2018Deep Co-Training for Semi-Supervised Image Recognitioncited by 147
  30. 2019Semi-Supervised Semantic Segmentation with High- and Low-level Consistencycited by 147
  31. 2021Semi-Supervised Semantic Segmentation with Pixel-Level Contrastive Learning from a Class-wise Memory Bankcited by 125
  32. 2020Guided Collaborative Training for Pixel-wise Semi-Supervised Learningcited by 120
  33. 2021Humble Teachers Teach Better Students for Semi-Supervised Object Detectioncited by 116
  34. 2021Perturbed and Strict Mean Teachers for Semi-supervised Semantic Segmentationcited by 113
  35. 2021Semi-supervised Semantic Segmentation with Directional Context-aware Consistencycited by 112
  36. 2021Pixel Contrastive-Consistent Semi-Supervised Semantic Segmentationcited by 101
  37. 2021Semi-Supervised Semantic Segmentation via Adaptive Equalization Learningcited by 101
  38. 2021Semi-Supervised Medical Image Segmentation via Cross Teaching between CNN and Transformercited by 100

50 most cited of 64 All 64 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 2; 2023: 25; 2024: 62; 2025: 59; 2026: 30.

178 papers.

Sort:

By area:

  1. 2024Depth Anything: Unleashing the Power of Large-Scale Unlabeled Datacited by 1,360
  2. 2024OpenVid-1M: A Large-Scale High-Quality Dataset for Text-to-video Generationcited by 341
  3. 2024OpenHumanVid: A Large-Scale High-Quality Dataset for Enhancing Human-Centric Video Generationcited by 77
  4. 2022Augmentation Matters: A Simple-yet-Effective Approach to Semi-supervised Semantic Segmentationcited by 64
  5. 2023Adapting Segment Anything Model for Change Detection in HR Remote Sensing Imagescited by 56
  6. 2024UniMatch V2: Pushing the Limit of Semi-Supervised Semantic Segmentationcited by 54
  7. 2023CorrMatch: Label Propagation via Correlation Matching for Semi-Supervised Semantic Segmentationcited by 44
  8. 2023SemiVL: Semi-Supervised Semantic Segmentation with Vision-Language Guidancecited by 41
  9. 2023VDD: Varied Drone Dataset for Semantic Segmentationcited by 40
  10. 2024Large-Scale 3D Medical Image Pre-training with Geometric Context Priorscited by 36
  11. 2023Conflict-Based Cross-View Consistency for Semi-Supervised Semantic Segmentationcited by 35
  12. 2024AllSpark: Reborn Labeled Features from Unlabeled in Transformer for Semi-Supervised Semantic Segmentationcited by 35
38 more
  1. 2024Adaptive Bidirectional Displacement for Semi-Supervised Medical Image Segmentationcited by 30
  2. 2024Audio-Visual Segmentation via Unlabeled Frame Exploitationcited by 28
  3. 2023Diverse Cotraining Makes Strong Semi-Supervised Segmentorcited by 26
  4. 2023Switching Temporary Teachers for Semi-Supervised Semantic Segmentationcited by 23
  5. 2024Learnable Prompting SAM-induced Knowledge Distillation for Semi-supervised Medical Image Segmentationcited by 23
  6. 2023Alternate Diverse Teaching for Semi-supervised Medical Image Segmentationcited by 22
  7. 2025SAMRefiner: Taming Segment Anything Model for Universal Mask Refinementcited by 20
  8. 2024SemiCD-VL: Visual-Language Model Guidance Makes Better Semi-supervised Change Detectorcited by 18
  9. 2024SAM as the Guide: Mastering Pseudo-Label Refinement in Semi-Supervised Referring Expression Segmentationcited by 17
  10. 2024Towards the Uncharted: Density-Descending Feature Perturbation for Semi-supervised Semantic Segmentationcited by 16
  11. 2024Cross Prompting Consistency with Segment Anything Model for Semi-supervised Medical Image Segmentationcited by 16
  12. 2024FreeTumor: Advance Tumor Segmentation via Large-Scale Tumor Synthesiscited by 15
  13. 2024DiffRect: Latent Diffusion Label Rectification for Semi-supervised Medical Image Segmentationcited by 15
  14. 2023All Points Matter: Entropy-Regularized Distribution Alignment for Weakly-supervised 3D Segmentationcited by 13
  15. 2023Jaccard Metric Losses: Optimizing the Jaccard Index with Soft Labelscited by 12
  16. 2024Semi-Supervised Semantic Segmentation Based on Pseudo-Labels: A Surveycited by 12
  17. 2024Beyond Pixels: Semi-Supervised Semantic Segmentation with a Multi-scale Patch-based Multi-Label Classifiercited by 12
  18. 2024CamoTeacher: Dual-Rotation Consistency Learning for Semi-Supervised Camouflaged Object Detectioncited by 12
  19. 2025Depth Anything at Any Conditioncited by 12
  20. 2024RESMatch: Referring Expression Segmentation in a Semi-Supervised Mannercited by 11
  21. 2025Semi-supervised Semantic Segmentation for Remote Sensing Images via Multi-scale Uncertainty Consistency and Cross-Teacher-Student Attentioncited by 10
  22. 2023Bridging Synthetic and Real Worlds for Pre-training Scene Text Detectorscited by 9
  23. 2025A Survey of Sample-Efficient Deep Learning for Change Detection in Remote Sensing: Tasks, Strategies, and Challengescited by 9
  24. 2023Semi-Supervised Semantic Segmentation via Marginal Contextual Informationcited by 8
  25. 2025Semi-supervised Semantic Segmentation with Multi-Constraint Consistency Learningcited by 8
  26. 20233D Medical Image Segmentation with Sparse Annotation via Cross-Teaching between 3D and 2D Networkscited by 7
  27. 2024C2F-SemiCD: A Coarse-to-Fine Semi-Supervised Change Detection Method Based on Consistency Regularization in High-Resolution Remote Sensing Imagescited by 7
  28. 2025DiffDoctor: Diagnosing Image Diffusion Models Before Treatingcited by 7
  29. 2025PoseTraj: Pose-Aware Trajectory Control in Video Diffusioncited by 7
  30. 2024Tuning-free Universally-Supervised Semantic Segmentationcited by 6
  31. 2024PrevMatch: Revisiting and Maximizing Temporal Knowledge in Semi-Supervised Semantic Segmentationcited by 6
  32. 2024Exploring Semi-Supervised Learning for Online Mappingcited by 6
  33. 2024Uncertainty-Participation Context Consistency Learning for Semi-supervised Semantic Segmentationcited by 6
  34. 2025When Confidence Fails: Revisiting Pseudo-Label Selection in Semi-supervised Semantic Segmentationcited by 6
  35. 2026DifferSeg: Towards Diverse Multimodal Binary Segmentation via Differential Perception and Frequency Guidancecited by 6
  36. 2023EC-Depth: Exploring the consistency of self-supervised monocular depth estimation in challenging scenescited by 5
  37. 2024PCT: Perspective Cue Training Framework for Multi-Camera BEV Segmentationcited by 5
  38. 2024ItTakesTwo: Leveraging Peer Representations for Semi-supervised LiDAR Semantic Segmentationcited by 5

50 most cited of 178 All 178 (JSON)