{"url":"/sota/unsupervised-action-segmentation-on-breakfast","task":{"name":"Unsupervised Action Segmentation","url":"/task/unsupervised-action-segmentation","note":null},"dataset":{"name":"Breakfast","url":"/dataset/breakfast"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"Unsupervised Action Segmentation is a challenging problem in high-level video understanding, where the goal is to segment a temporally untrimmed sequence into distinct action segments without access to ground truth labels during training. Unlike supervised methods, which rely on annotated datasets, unsupervised approaches aim to discover the underlying structure of actions directly from data. This makes the task particularly valuable for scenarios with limited labeled data or large-scale unlabeled video datasets. The results of Unsupervised Action Segmentation can be further applied to tasks such as action localization and video summarization.","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":["F1","Acc","JSD","Precision","Recall","mIoU"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"F1":"higher","Acc":"higher","JSD":null,"Precision":"higher","Recall":"higher","mIoU":null}},"counts":{"rows":8,"rows_with_code":7,"rows_with_paper_page":8,"rows_dated":8,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"HVQ","metrics":{"Acc":"54.4","F1":"39.7","JSD":"82.5","Precision":"35.6","Recall":"44.9"},"uses_additional_data":false,"paper_date":"2024-12-23","paper":"/paper/hierarchical-vector-quantization-for","paper_url":"https://arxiv.org/abs/2412.17640v2","paper_title":"Hierarchical Vector Quantization for Unsupervised Action Segmentation","code":"https://github.com/fedespu/hvq","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"ASOT","metrics":{"Acc":"56.1","F1":"38.3","JSD":"94.9","Precision":"36.7","Recall":"40.1","mIoU":"18.6"},"uses_additional_data":false,"paper_date":"2024-04-01","paper":"/paper/temporally-consistent-unbalanced-optimal","paper_url":"https://arxiv.org/abs/2404.01518v3","paper_title":"Temporally Consistent Unbalanced Optimal Transport for Unsupervised Action Segmentation","code":"https://github.com/mingu6/action_seg_ot","n_code_links":1,"syntology":null},{"rank_in_archive_order":3,"model":"UFSA","metrics":{"Acc":"52.1","F1":"38"},"uses_additional_data":false,"paper_date":"2023-05-31","paper":"/paper/permutation-aware-action-segmentation-via","paper_url":"https://arxiv.org/abs/2305.19478v4","paper_title":"Permutation-Aware Action Segmentation via Unsupervised Frame-to-Segment Alignment","code":"https://github.com/trquhuytin/TOT-CVPR22","n_code_links":1,"syntology":null},{"rank_in_archive_order":4,"model":"ASAL","metrics":{"Acc":"52.5","F1":"37.9"},"uses_additional_data":false,"paper_date":"2021-04-05","paper":"/paper/action-shuffle-alternating-learning-for","paper_url":"https://arxiv.org/abs/2104.02116v1","paper_title":"Action Shuffle Alternating Learning for Unsupervised Action Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":5,"model":"TOT","metrics":{"Acc":"47.5","F1":"31.0","JSD":"90.2","Precision":"37.7","Recall":"26.3"},"uses_additional_data":false,"paper_date":"2021-05-27","paper":"/paper/unsupervised-activity-segmentation-by-joint","paper_url":"https://arxiv.org/abs/2105.13353v7","paper_title":"Unsupervised Action Segmentation by Joint Representation Learning and Online Clustering","code":"https://github.com/trquhuytin/TOT-CVPR22","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":2,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":6,"model":"TOT+TCL","metrics":{"Acc":"39.0","F1":"30.3","JSD":"85.6","Precision":"26.2","Recall":"36.0"},"uses_additional_data":false,"paper_date":"2021-05-27","paper":"/paper/unsupervised-activity-segmentation-by-joint","paper_url":"https://arxiv.org/abs/2105.13353v7","paper_title":"Unsupervised Action Segmentation by Joint Representation Learning and Online Clustering","code":"https://github.com/trquhuytin/TOT-CVPR22","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":2,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":7,"model":"CTE","metrics":{"Acc":"41.8","F1":"26.4","JSD":"87.4","Precision":"25.8","Recall":"27.0"},"uses_additional_data":false,"paper_date":"2019-04-08","paper":"/paper/unsupervised-learning-of-action-classes-with","paper_url":"http://arxiv.org/abs/1904.04189v1","paper_title":"Unsupervised learning of action classes with continuous temporal embedding","code":"https://github.com/annusha/unsup_temp_embed","n_code_links":2,"syntology":null},{"rank_in_archive_order":8,"model":"LSTM+AL","metrics":{"Acc":"42.9","mIoU":"46.9"},"uses_additional_data":false,"paper_date":"2018-11-12","paper":"/paper/a-perceptual-prediction-framework-for-self","paper_url":"http://arxiv.org/abs/1811.04869v3","paper_title":"A Perceptual Prediction Framework for Self Supervised Event Segmentation","code":"https://github.com/CVPRUSFTampa/EventSegmentation","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}}],"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 6,885 of the 9,623 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":9623,"papers_checked":6885,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":2737},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-25T09:33:49+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":3,"rows_with_any_sample_ran":0,"distinct_papers_with_graph_line":2,"distinct_papers_with_any_sample_ran":0,"samples_over_distinct_papers":{"n_ran":0,"n_unverified":3,"n_samples":3,"n_pointer_only_licence":0,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":0,"n_unverified":5,"n_samples":5,"n_pointer_only_licence":0,"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"}}}