{"url":"/sota/semi-supervised-semantic-segmentation-on-4","task":{"name":"Semi-Supervised Semantic Segmentation","url":"/task/semi-supervised-semantic-segmentation","note":null},"dataset":{"name":"Pascal VOC 2012 12.5% labeled","url":null},"category":"Computer Vision","categories":["Computer Code","Computer Vision","Medical","Robots"],"category_note":null,"description":"Models that are trained with a small number of labeled examples and a large number of unlabeled examples and whose aim is to learn to segment an image (i.e. assign a class to every pixel).","description_from":"task","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["Validation mIoU"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Validation mIoU":null}},"counts":{"rows":38,"rows_with_code":35,"rows_with_paper_page":38,"rows_dated":38,"rows_using_additional_data":12},"rows":[{"rank_in_archive_order":1,"model":"SemiOVS (w/ UniMatch, ResNet-101)","metrics":{"Validation mIoU":"82.4"},"uses_additional_data":true,"paper_date":"2025-07-04","paper":"/paper/leveraging-out-of-distribution-unlabeled","paper_url":"https://arxiv.org/abs/2507.03302","paper_title":"Leveraging Out-of-Distribution Unlabeled Images: Semi-Supervised Semantic Segmentation with an Open-Vocabulary Model","code":"https://github.com/wooseok-shin/SemiOVS","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"SemiOVS (w/ PrevMatch, ResNet-101)","metrics":{"Validation mIoU":"82.3"},"uses_additional_data":true,"paper_date":"2025-07-04","paper":"/paper/leveraging-out-of-distribution-unlabeled","paper_url":"https://arxiv.org/abs/2507.03302","paper_title":"Leveraging Out-of-Distribution Unlabeled Images: Semi-Supervised Semantic Segmentation with an Open-Vocabulary Model","code":"https://github.com/wooseok-shin/SemiOVS","n_code_links":1,"syntology":null},{"rank_in_archive_order":3,"model":"AllSpark","metrics":{"Validation mIoU":"82.04%"},"uses_additional_data":false,"paper_date":"2024-03-04","paper":"/paper/allspark-reborn-labeled-features-from","paper_url":"https://arxiv.org/abs/2403.01818v3","paper_title":"AllSpark: Reborn Labeled Features from Unlabeled in Transformer for Semi-Supervised Semantic Segmentation","code":"https://github.com/xmed-lab/AllSpark","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":1,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":4,"model":"UniMatch","metrics":{"Validation mIoU":"81.92%"},"uses_additional_data":false,"paper_date":"2022-08-21","paper":"/paper/revisiting-weak-to-strong-consistency-in-semi","paper_url":"https://arxiv.org/abs/2208.09910v2","paper_title":"Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentation","code":"https://github.com/LiheYoung/UniMatch","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":5,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":5,"model":"PrevMatch (ResNet-101)","metrics":{"Validation mIoU":"81.9"},"uses_additional_data":false,"paper_date":"2024-05-31","paper":"/paper/revisiting-and-maximizing-temporal-knowledge","paper_url":"https://arxiv.org/abs/2405.20610v1","paper_title":"Revisiting and Maximizing Temporal Knowledge in Semi-supervised Semantic Segmentation","code":"https://github.com/wooseok-shin/PrevMatch","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":5,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":6,"model":"CorrMatch (Deeplabv3+ with ResNet-101)","metrics":{"Validation mIoU":"81.9%"},"uses_additional_data":false,"paper_date":"2023-06-07","paper":"/paper/corrmatch-label-propagation-via-correlation","paper_url":"https://arxiv.org/abs/2306.04300v3","paper_title":"CorrMatch: Label Propagation via Correlation Matching for Semi-Supervised Semantic Segmentation","code":"https://github.com/bbbbchan/corrmatch","n_code_links":1,"syntology":null},{"rank_in_archive_order":7,"model":"Dual Teacher","metrics":{"Validation mIoU":"81.19"},"uses_additional_data":false,"paper_date":"2023-09-21","paper":"/paper/switching-temporary-teachers-for-semi","paper_url":"https://openreview.net/forum?id=JXvszuOqY3","paper_title":"Switching Temporary Teachers for Semi-Supervised Semantic Segmentation","code":"https://github.com/naver-ai/dual-teacher","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"PCR (DeepLab v3+ with ResNet-101 pretraind on ImageNet-1K)","metrics":{"Validation mIoU":"80.71%"},"uses_additional_data":false,"paper_date":"2022-10-10","paper":"/paper/semi-supervised-semantic-segmentation-with-6","paper_url":"https://arxiv.org/abs/2210.04388v1","paper_title":"Semi-supervised Semantic Segmentation with Prototype-based Consistency Regularization","code":"https://github.com/heimingx/semi_seg_proto","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"S4MC","metrics":{"Validation mIoU":"79.67%"},"uses_additional_data":false,"paper_date":"2023-08-26","paper":"/paper/semi-supervised-semantic-segmentation-via-3","paper_url":"https://arxiv.org/abs/2308.13900v2","paper_title":"Semi-Supervised Semantic Segmentation via Marginal Contextual Information","code":"https://github.com/s4mcontext/s4mc","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"U2PL (DeepLab v3+ with ResNet-101 pretraind on ImageNet-1K, CutMix)","metrics":{"Validation mIoU":"79.01%"},"uses_additional_data":false,"paper_date":"2022-03-08","paper":"/paper/semi-supervised-semantic-segmentation-using-2","paper_url":"https://arxiv.org/abs/2203.03884v2","paper_title":"Semi-Supervised Semantic Segmentation Using Unreliable Pseudo-Labels","code":"https://github.com/Haochen-Wang409/U2PL","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":11,"model":"PS-MT","metrics":{"Validation mIoU":"78.20%"},"uses_additional_data":false,"paper_date":"2021-11-25","paper":"/paper/perturbed-and-strict-mean-teachers-for-semi","paper_url":"https://arxiv.org/abs/2111.12903v3","paper_title":"Perturbed and Strict Mean Teachers for Semi-supervised Semantic Segmentation","code":"https://github.com/yyliu01/ps-mt","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"FARCLUSS","metrics":{"Validation mIoU":"78.2"},"uses_additional_data":false,"paper_date":"2025-06-11","paper":"/paper/farcluss-fuzzy-adaptive-rebalancing-and","paper_url":"https://arxiv.org/abs/2506.11142v2","paper_title":"FARCLUSS: Fuzzy Adaptive Rebalancing and Contrastive Uncertainty Learning for Semi-Supervised Semantic Segmentation","code":"https://github.com/psychofict/FARCLUSS","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"n-CPS (ResNet-101)","metrics":{"Validation mIoU":"77.99%"},"uses_additional_data":false,"paper_date":"2021-12-14","paper":"/paper/n-cps-generalising-cross-pseudo-supervision","paper_url":"https://arxiv.org/abs/2112.07528v4","paper_title":"n-CPS: Generalising Cross Pseudo Supervision to n Networks for Semi-Supervised Semantic Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":14,"model":"AEL (DeepLab v3+ with ResNet-101 pretraind on ImageNet-1K)","metrics":{"Validation mIoU":"77.57%"},"uses_additional_data":false,"paper_date":"2021-10-11","paper":"/paper/semi-supervised-semantic-segmentation-via-2","paper_url":"https://arxiv.org/abs/2110.05474v1","paper_title":"Semi-Supervised Semantic Segmentation via Adaptive Equalization Learning","code":"https://github.com/hzhupku/semiseg-ael","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":15,"model":"CPS","metrics":{"Validation mIoU":"76.44%"},"uses_additional_data":false,"paper_date":"2021-06-02","paper":"/paper/semi-supervised-semantic-segmentation-with-3","paper_url":"https://arxiv.org/abs/2106.01226v2","paper_title":"Semi-Supervised Semantic Segmentation with Cross Pseudo Supervision","code":"https://github.com/charlesCXK/TorchSemiSeg","n_code_links":3,"syntology":{"n_ran":22,"n_unverified":17,"n_samples":39,"n_pointer_only_licence":28}},{"rank_in_archive_order":16,"model":"GuidedMix-Net","metrics":{"Validation mIoU":"76.4%"},"uses_additional_data":false,"paper_date":"2021-06-29","paper":"/paper/guidedmix-net-learning-to-improve-pseudo","paper_url":"https://arxiv.org/abs/2106.15064v2","paper_title":"GuidedMix-Net: Learning to Improve Pseudo Masks Using Labeled Images as Reference","code":"https://github.com/yh-pengtu/GuidedMix-Net","n_code_links":1,"syntology":null},{"rank_in_archive_order":17,"model":"CPCL (DeepLab v3+ with ResNet-101)","metrics":{"Validation mIoU":"76.4%"},"uses_additional_data":false,"paper_date":"2022-11-30","paper":"/paper/conservative-progressive-collaborative","paper_url":"https://arxiv.org/abs/2211.16701v2","paper_title":"Conservative-Progressive Collaborative Learning for Semi-supervised Semantic Segmentation","code":"https://github.com/leofansq/CPCL","n_code_links":1,"syntology":null},{"rank_in_archive_order":18,"model":"CW-BASS (DeepLab v3+ with ResNet-50)","metrics":{"Validation mIoU":"75.81%"},"uses_additional_data":false,"paper_date":"2025-02-21","paper":"/paper/confidence-weighted-boundary-aware-learning","paper_url":"https://arxiv.org/abs/2502.15152v1","paper_title":"Confidence-Weighted Boundary-Aware Learning for Semi-Supervised Semantic Segmentation","code":"https://github.com/psychofict/CW-BASS","n_code_links":1,"syntology":null},{"rank_in_archive_order":19,"model":"PS-MT","metrics":{"Validation mIoU":"75.70%"},"uses_additional_data":false,"paper_date":"2021-11-25","paper":"/paper/perturbed-and-strict-mean-teachers-for-semi","paper_url":"https://arxiv.org/abs/2111.12903v3","paper_title":"Perturbed and Strict Mean Teachers for Semi-supervised Semantic Segmentation","code":"https://github.com/yyliu01/ps-mt","n_code_links":1,"syntology":null},{"rank_in_archive_order":20,"model":"PCT (DeepLab v3+ with ResNet-50 pretrained on ImageNet-1K)","metrics":{"Validation mIoU":"75.52%"},"uses_additional_data":false,"paper_date":"2022-08-02","paper":"/paper/learning-pseudo-labels-for-semi-and-weakly","paper_url":"https://www.sciencedirect.com/science/article/pii/S003132032200406X","paper_title":"Learning Pseudo Labels for Semi-and-Weakly Supervised Semantic Segmentation","code":"https://github.com/YudeWang/Learning-Pseudo-Label","n_code_links":1,"syntology":null},{"rank_in_archive_order":21,"model":"Error Localization Network","metrics":{"Validation mIoU":"75.10%"},"uses_additional_data":true,"paper_date":"2022-04-05","paper":"/paper/semi-supervised-semantic-segmentation-with-5","paper_url":"https://arxiv.org/abs/2204.02078v3","paper_title":"Semi-supervised Semantic Segmentation with Error Localization Network","code":"https://github.com/kinux98/SSL_ELN","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":6,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":22,"model":"n-CPS","metrics":{"Validation mIoU":"74.21%"},"uses_additional_data":false,"paper_date":"2021-12-14","paper":"/paper/n-cps-generalising-cross-pseudo-supervision","paper_url":"https://arxiv.org/abs/2112.07528v4","paper_title":"n-CPS: Generalising Cross Pseudo Supervision to n Networks for Semi-Supervised Semantic Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":23,"model":"CPCL (DeepLab v3+ with ResNet-50)","metrics":{"Validation mIoU":"73.74%"},"uses_additional_data":false,"paper_date":"2022-11-30","paper":"/paper/conservative-progressive-collaborative","paper_url":"https://arxiv.org/abs/2211.16701v2","paper_title":"Conservative-Progressive Collaborative Learning for Semi-supervised Semantic Segmentation","code":"https://github.com/leofansq/CPCL","n_code_links":1,"syntology":null},{"rank_in_archive_order":24,"model":"GuidedMix-Net","metrics":{"Validation mIoU":"73.4%"},"uses_additional_data":true,"paper_date":"2021-06-29","paper":"/paper/guidedmix-net-learning-to-improve-pseudo","paper_url":"https://arxiv.org/abs/2106.15064v2","paper_title":"GuidedMix-Net: Learning to Improve Pseudo Masks Using Labeled Images as Reference","code":"https://github.com/yh-pengtu/GuidedMix-Net","n_code_links":1,"syntology":null},{"rank_in_archive_order":25,"model":"Error Localization Network","metrics":{"Validation mIoU":"73.2%"},"uses_additional_data":true,"paper_date":"2022-04-05","paper":"/paper/semi-supervised-semantic-segmentation-with-5","paper_url":"https://arxiv.org/abs/2204.02078v3","paper_title":"Semi-supervised Semantic Segmentation with Error Localization Network","code":"https://github.com/kinux98/SSL_ELN","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":6,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":26,"model":"DMT","metrics":{"Validation mIoU":"72.70%"},"uses_additional_data":false,"paper_date":"2020-04-18","paper":"/paper/semi-supervised-semantic-segmentation-via","paper_url":"https://arxiv.org/abs/2004.08514v4","paper_title":"DMT: Dynamic Mutual Training for Semi-Supervised Learning","code":"https://github.com/voldemortX/DST-CBC","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":27,"model":"CutMix","metrics":{"Validation mIoU":"72.45%"},"uses_additional_data":true,"paper_date":"2019-06-05","paper":"/paper/consistency-regularization-and-cutmix-for","paper_url":"https://arxiv.org/abs/1906.01916v5","paper_title":"Semi-supervised semantic segmentation needs strong, varied perturbations","code":"https://github.com/ZHKKKe/PixelSSL","n_code_links":5,"syntology":{"n_ran":0,"n_unverified":3,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":28,"model":"SemiSegContrast","metrics":{"Validation mIoU":"71.6%"},"uses_additional_data":false,"paper_date":"2021-04-27","paper":"/paper/semi-supervised-semantic-segmentation-with-2","paper_url":"https://arxiv.org/abs/2104.13415v3","paper_title":"Semi-Supervised Semantic Segmentation with Pixel-Level Contrastive Learning from a Class-wise Memory Bank","code":"https://github.com/Shathe/SemiSeg-Contrastive","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":3,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":29,"model":"s4GAN + MLMT","metrics":{"Validation mIoU":"71.4%"},"uses_additional_data":true,"paper_date":"2019-08-15","paper":"/paper/semi-supervised-semantic-segmentation-with","paper_url":"https://arxiv.org/abs/1908.05724v1","paper_title":"Semi-Supervised Semantic Segmentation with High- and Low-level Consistency","code":"https://github.com/sud0301/semisup-semseg","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":0,"n_samples":4,"n_pointer_only_licence":4}},{"rank_in_archive_order":30,"model":"ClassMix","metrics":{"Validation mIoU":"71.00%"},"uses_additional_data":false,"paper_date":"2020-07-15","paper":"/paper/classmix-segmentation-based-data-augmentation","paper_url":"https://arxiv.org/abs/2007.07936v2","paper_title":"ClassMix: Segmentation-Based Data Augmentation for Semi-Supervised Learning","code":"https://github.com/lorenmt/reco","n_code_links":2,"syntology":{"n_ran":5,"n_unverified":14,"n_samples":19,"n_pointer_only_licence":0}},{"rank_in_archive_order":31,"model":"ReCo","metrics":{"Validation mIoU":"71.00%"},"uses_additional_data":true,"paper_date":"2021-04-09","paper":"/paper/bootstrapping-semantic-segmentation-with","paper_url":"https://arxiv.org/abs/2104.04465v4","paper_title":"Bootstrapping Semantic Segmentation with Regional Contrast","code":"https://github.com/lorenmt/reco","n_code_links":2,"syntology":{"n_ran":3,"n_unverified":3,"n_samples":6,"n_pointer_only_licence":2}},{"rank_in_archive_order":32,"model":"GIST and RIST","metrics":{"Validation mIoU":"70.76%"},"uses_additional_data":false,"paper_date":"2021-03-31","paper":"/paper/the-gist-and-rist-of-iterative-self-training","paper_url":"https://arxiv.org/abs/2103.17105v3","paper_title":"The GIST and RIST of Iterative Self-Training for Semi-Supervised Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":33,"model":"s4GAN+MLMT","metrics":{"Validation mIoU":"70.4%"},"uses_additional_data":true,"paper_date":"2019-08-15","paper":"/paper/semi-supervised-semantic-segmentation-with","paper_url":"https://arxiv.org/abs/1908.05724v1","paper_title":"Semi-Supervised Semantic Segmentation with High- and Low-level Consistency","code":"https://github.com/sud0301/semisup-semseg","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":0,"n_samples":4,"n_pointer_only_licence":4}},{"rank_in_archive_order":34,"model":"CutMix","metrics":{"Validation mIoU":"67.6%"},"uses_additional_data":true,"paper_date":"2019-06-05","paper":"/paper/consistency-regularization-and-cutmix-for","paper_url":"https://arxiv.org/abs/1906.01916v5","paper_title":"Semi-supervised semantic segmentation needs strong, varied perturbations","code":"https://github.com/ZHKKKe/PixelSSL","n_code_links":5,"syntology":{"n_ran":0,"n_unverified":3,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":35,"model":"s4GAN+MLMT","metrics":{"Validation mIoU":"67.3%"},"uses_additional_data":true,"paper_date":"2019-08-15","paper":"/paper/semi-supervised-semantic-segmentation-with","paper_url":"https://arxiv.org/abs/1908.05724v1","paper_title":"Semi-Supervised Semantic Segmentation with High- and Low-level Consistency","code":"https://github.com/sud0301/semisup-semseg","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":0,"n_samples":4,"n_pointer_only_licence":4}},{"rank_in_archive_order":36,"model":"Dense FixMatch (DeepLabv3+ ResNet-101, over-sampling, single pass eval)","metrics":{"Validation mIoU":"65.82%"},"uses_additional_data":false,"paper_date":"2022-10-18","paper":"/paper/dense-fixmatch-a-simple-semi-supervised","paper_url":"https://arxiv.org/abs/2210.09919v1","paper_title":"Dense FixMatch: a simple semi-supervised learning method for pixel-wise prediction tasks","code":"https://github.com/miquelmarti/DenseFixMatch","n_code_links":1,"syntology":null},{"rank_in_archive_order":37,"model":"Adversarial","metrics":{"Validation mIoU":"64.3%"},"uses_additional_data":true,"paper_date":"2018-02-22","paper":"/paper/adversarial-learning-for-semi-supervised","paper_url":"http://arxiv.org/abs/1802.07934v2","paper_title":"Adversarial Learning for Semi-Supervised Semantic Segmentation","code":"https://github.com/wasidennis/AdaptSegNet","n_code_links":13,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":38,"model":"Dense FixMatch (DeepLabv3+ ResNet-50, over-sampling, single pass eval)","metrics":{"Validation mIoU":"62.49%"},"uses_additional_data":false,"paper_date":"2022-10-18","paper":"/paper/dense-fixmatch-a-simple-semi-supervised","paper_url":"https://arxiv.org/abs/2210.09919v1","paper_title":"Dense FixMatch: a simple semi-supervised learning method for pixel-wise prediction tasks","code":"https://github.com/miquelmarti/DenseFixMatch","n_code_links":1,"syntology":null}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 6,264 of the 9,581 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9581,"papers_checked":6264,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":3316},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-24T18:15:14+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":18,"rows_with_any_sample_ran":16,"distinct_papers_with_graph_line":14,"distinct_papers_with_any_sample_ran":13,"samples_over_distinct_papers":{"n_ran":51,"n_unverified":57,"n_samples":108,"n_pointer_only_licence":35,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":60,"n_unverified":66,"n_samples":126,"n_pointer_only_licence":43,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}