{"url":"/sota/audio-deepfake-detection-on-asvspoof-2021","task":{"name":"Audio Deepfake Detection","url":"/task/audio-deepfake-detection","note":null},"dataset":{"name":"ASVspoof 2021","url":"/dataset/asvspoof-2021"},"category":"Computer Vision","categories":["Audio","Computer Vision","Miscellaneous","Speech"],"category_note":null,"description":"Nowadays, deepfake is now generically used by the media\r\nor people to refer to any audio or video in which important\r\nattributes have been either digitally altered or swapped,\r\nwith the help of artificial intelligence (AI). Audio deepfake detection is a task that aims to distinguish genuine utterances from fake ones via machine\r\nlearning techniques.","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":["21LA EER","21DF EER"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"21LA EER":null,"21DF EER":null}},"counts":{"rows":8,"rows_with_code":8,"rows_with_paper_page":8,"rows_dated":8,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"XLSR-Mamba","metrics":{"21DF EER":"1.88","21LA EER":"0.93"},"uses_additional_data":false,"paper_date":"2024-11-15","paper":"/paper/xlsr-mamba-a-dual-column-bidirectional-state","paper_url":"https://arxiv.org/abs/2411.10027v2","paper_title":"XLSR-Mamba: A Dual-Column Bidirectional State Space Model for Spoofing Attack Detection","code":"https://github.com/swagshaw/xlsr-mamba","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"XLSR+AASIST","metrics":{"21DF EER":"3.69","21LA EER":"1.0"},"uses_additional_data":false,"paper_date":"2022-02-24","paper":"/paper/automatic-speaker-verification-spoofing-and","paper_url":"https://arxiv.org/abs/2202.12233v2","paper_title":"Automatic speaker verification spoofing and deepfake detection using wav2vec 2.0 and data augmentation","code":"https://github.com/takhemlata/ssl_anti-spoofing","n_code_links":3,"syntology":{"n_ran":1,"n_unverified":1,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":3,"model":"XLSR+SLS","metrics":{"21DF EER":"1.96","21LA EER":"2.86"},"uses_additional_data":false,"paper_date":"2024-10-28","paper":"/paper/audio-deepfake-detection-with-self-supervised-1","paper_url":"https://openreview.net/pdf?id=acJMIXJg2u","paper_title":"Audio Deepfake Detection with Self-Supervised XLS-R and SLS Classifier","code":"https://github.com/qishanzhang/slsforadd","n_code_links":2,"syntology":null},{"rank_in_archive_order":4,"model":"TCM-Add","metrics":{"21DF EER":"2.14","21LA EER":"2.99"},"uses_additional_data":false,"paper_date":"2024-06-25","paper":"/paper/temporal-channel-modeling-in-multi-head-self","paper_url":"https://arxiv.org/abs/2406.17376v1","paper_title":"Temporal-Channel Modeling in Multi-head Self-Attention for Synthetic Speech Detection","code":"https://github.com/ductuantruong/tcm_add","n_code_links":1,"syntology":null},{"rank_in_archive_order":5,"model":"BTS-E","metrics":{"21DF EER":"/","21LA EER":"8.75"},"uses_additional_data":false,"paper_date":"2023-05-05","paper":"/paper/bts-e-audio-deepfake-detection-using","paper_url":"https://ieeexplore.ieee.org/abstract/document/10095927","paper_title":"Bts-e: Audio deepfake detection using breathing-talking-silence encoder","code":"https://github.com/josebeo2016/BTS-Encoder-ASVspoof","n_code_links":1,"syntology":null},{"rank_in_archive_order":6,"model":"RawGAT-ST","metrics":{"21DF EER":"23.26","21LA EER":"10.25"},"uses_additional_data":false,"paper_date":"2021-07-27","paper":"/paper/end-to-end-spectro-temporal-graph-attention","paper_url":"https://arxiv.org/abs/2107.12710v2","paper_title":"End-to-End Spectro-Temporal Graph Attention Networks for Speaker Verification Anti-Spoofing and Speech Deepfake Detection","code":"https://github.com/eurecom-asp/rawgat-st-antispoofing","n_code_links":1,"syntology":null},{"rank_in_archive_order":7,"model":"AASIST","metrics":{"21DF EER":"21.07","21LA EER":"11.46"},"uses_additional_data":false,"paper_date":"2021-10-04","paper":"/paper/aasist-audio-anti-spoofing-using-integrated","paper_url":"https://arxiv.org/abs/2110.01200v1","paper_title":"AASIST: Audio Anti-Spoofing using Integrated Spectro-Temporal Graph Attention Networks","code":"https://github.com/clovaai/aasist","n_code_links":2,"syntology":{"n_ran":11,"n_unverified":1,"n_samples":12,"n_pointer_only_licence":0}},{"rank_in_archive_order":8,"model":"RawNet-2","metrics":{"21DF EER":"40.06","21LA EER":"40.07"},"uses_additional_data":false,"paper_date":"2020-11-02","paper":"/paper/end-to-end-anti-spoofing-with-rawnet2","paper_url":"https://arxiv.org/abs/2011.01108v3","paper_title":"End-to-end anti-spoofing with RawNet2","code":"https://github.com/eurecom-asp/rawnet2-antispoofing","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"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 6,795 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":6795,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":2785},"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":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":15,"n_unverified":2,"n_samples":17,"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":15,"n_unverified":2,"n_samples":17,"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"}}}