{"url":"/sota/sound-event-detection-on-desed","task":{"name":"Sound Event Detection","url":"/task/sound-event-detection","note":null},"dataset":{"name":"DESED","url":"/dataset/desed"},"category":"Audio","categories":["Audio"],"category_note":null,"description":"**Sound Event Detection** (SED) is the task of recognizing the sound events and their respective temporal start and end time in a recording. Sound events in real life do not always occur in isolation, but tend to considerably overlap with each other. Recognizing such overlapping sound events is referred as polyphonic SED.\n\n\n<span class=\"description-source\">Source: [A report on sound event detection with different binaural features ](https://arxiv.org/abs/1710.02997)</span>","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":["event-based F1 score","PSDS1","PSDS2"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"event-based F1 score":"higher","PSDS1":null,"PSDS2":null}},"counts":{"rows":13,"rows_with_code":12,"rows_with_paper_page":13,"rows_dated":13,"rows_using_additional_data":5},"rows":[{"rank_in_archive_order":1,"model":"ATST-SED","metrics":{"PSDS1":"0.583","PSDS2":"0.810","event-based F1 score":"63.4"},"uses_additional_data":true,"paper_date":"2023-09-15","paper":"/paper/fine-tune-the-pretrained-atst-model-for-sound","paper_url":"https://arxiv.org/abs/2309.08153v2","paper_title":"Fine-tune the pretrained ATST model for sound event detection","code":"https://github.com/Audio-WestlakeU/ATST-SED","n_code_links":1,"syntology":{"n_ran":10,"n_unverified":0,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":2,"model":"SE-CRNN-16 with DualKD","metrics":{"PSDS1":"0.474","PSDS2":"0.698","event-based F1 score":"55.6"},"uses_additional_data":true,"paper_date":"2024-02-05","paper":"/paper/dual-knowledge-distillation-for-efficient","paper_url":"https://arxiv.org/abs/2402.02781v1","paper_title":"Dual Knowledge Distillation for Efficient Sound Event Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":3,"model":"FDY-CRNN","metrics":{"PSDS1":"0.452","PSDS2":"0.670","event-based F1 score":"54.0"},"uses_additional_data":false,"paper_date":"2022-03-29","paper":"/paper/frequency-dynamic-convolution-frequency","paper_url":"https://arxiv.org/abs/2203.15296v2","paper_title":"Frequency Dynamic Convolution: Frequency-Adaptive Pattern Recognition for Sound Event Detection","code":"https://github.com/frednam93/FDY-SED","n_code_links":1,"syntology":{"n_ran":8,"n_unverified":3,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":4,"model":"HTS-AT","metrics":{"event-based F1 score":"50.7"},"uses_additional_data":false,"paper_date":"2022-02-02","paper":"/paper/hts-at-a-hierarchical-token-semantic-audio","paper_url":"https://arxiv.org/abs/2202.00874v1","paper_title":"HTS-AT: A Hierarchical Token-Semantic Audio Transformer for Sound Classification and Detection","code":"https://github.com/retrocirce/hts-audio-transformer","n_code_links":1,"syntology":{"n_ran":7,"n_unverified":2,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":5,"model":"RCT","metrics":{"PSDS1":"0.4395","PSDS2":"0.6711","event-based F1 score":"49.62"},"uses_additional_data":false,"paper_date":"2021-10-21","paper":"/paper/rct-random-consistency-training-for-semi","paper_url":"https://arxiv.org/abs/2110.11144v3","paper_title":"RCT: Random Consistency Training for Semi-supervised Sound Event Detection","code":"https://github.com/Audio-WestlakeU/RCT-Random-Consistency-Training","n_code_links":2,"syntology":null},{"rank_in_archive_order":6,"model":"FiltAug SED","metrics":{"PSDS1":"0.4336","PSDS2":"0.8161","event-based F1 score":"49.6"},"uses_additional_data":false,"paper_date":"2021-07-08","paper":"/paper/heavily-augmented-sound-event-detection","paper_url":"https://arxiv.org/abs/2107.03649v3","paper_title":"Heavily Augmented Sound Event Detection utilizing Weak Predictions","code":"https://github.com/frednam93/FilterAugSED","n_code_links":1,"syntology":null},{"rank_in_archive_order":7,"model":"SED-SSep baseline dcase task 4 2020 v2","metrics":{"event-based F1 score":"40.7"},"uses_additional_data":false,"paper_date":"2020-07-08","paper":"/paper/improving-sound-event-detection-in-domestic-1","paper_url":"https://arxiv.org/abs/2007.03932v1","paper_title":"Improving Sound Event Detection In Domestic Environments Using Sound Separation","code":"https://github.com/turpaultn/dcase20_task4","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"Baseline dcase task 4 2020 v2","metrics":{"event-based F1 score":"39.0"},"uses_additional_data":false,"paper_date":"2020-07-08","paper":"/paper/training-sound-event-detection-on-a-1","paper_url":"https://arxiv.org/abs/2007.03931v1","paper_title":"Training Sound Event Detection On A Heterogeneous Dataset","code":"https://github.com/turpaultn/dcase20_task4","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"Baseline","metrics":{"event-based F1 score":"25.8"},"uses_additional_data":false,"paper_date":"2019-10-26","paper":"/paper/sound-event-detection-in-domestic","paper_url":"https://hal.inria.fr/hal-02160855","paper_title":"Sound event detection in domestic environments withweakly labeled data and soundscape synthesis","code":"https://github.com/turpaultn/DCASE2019_task4","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"MAT-SED","metrics":{"PSDS1":"0.587","PSDS2":"0.896"},"uses_additional_data":true,"paper_date":"2024-08-16","paper":"/paper/mat-sed-amasked-audio-transformer-with-masked","paper_url":"https://arxiv.org/abs/2408.08673v2","paper_title":"MAT-SED: A Masked Audio Transformer with Masked-Reconstruction Based Pre-training for Sound Event Detection","code":"https://github.com/cai525/transformer4sed","n_code_links":1,"syntology":null},{"rank_in_archive_order":11,"model":"JiTTER","metrics":{"PSDS1":"0.584"},"uses_additional_data":true,"paper_date":"2025-02-28","paper":"/paper/jitter-jigsaw-temporal-transformer-for-event","paper_url":"https://arxiv.org/abs/2502.20857v1","paper_title":"JiTTER: Jigsaw Temporal Transformer for Event Reconstruction for Self-Supervised Sound Event Detection","code":"https://github.com/frednam93/JiTTER-SED","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"ABC + MDFD-CRNN","metrics":{"PSDS1":"0.577"},"uses_additional_data":true,"paper_date":"2024-06-19","paper":"/paper/pushing-the-limit-of-sound-event-detection","paper_url":"https://arxiv.org/abs/2406.13312v3","paper_title":"Pushing the Limit of Sound Event Detection with Multi-Dilated Frequency Dynamic Convolution","code":"https://github.com/frednam93/MDFD-SED","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"MDFD-CRNN","metrics":{"PSDS1":"0.485"},"uses_additional_data":false,"paper_date":"2024-06-19","paper":"/paper/pushing-the-limit-of-sound-event-detection","paper_url":"https://arxiv.org/abs/2406.13312v3","paper_title":"Pushing the Limit of Sound Event Detection with Multi-Dilated Frequency Dynamic Convolution","code":"https://github.com/frednam93/MDFD-SED","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 6,885 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":3,"distinct_papers_with_graph_line":3,"distinct_papers_with_any_sample_ran":3,"samples_over_distinct_papers":{"n_ran":25,"n_unverified":5,"n_samples":30,"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":25,"n_unverified":5,"n_samples":30,"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"}}}