{"url":"/dataset/wenetspeech","name":"WenetSpeech","full_name":null,"description_markdown":"WenetSpeech is a multi-domain Mandarin corpus consisting of 10,000+ hours high-quality labeled speech, 2,400+ hours weakly labelled speech, and about 10,000 hours unlabeled speech, with 22,400+ hours in total. The authors collected the data from YouTube and Podcast, which covers a variety of speaking styles, scenarios, domains, topics, and noisy conditions. An optical character recognition (OCR) based method is introduced to generate the audio/text segmentation candidates for the YouTube data on its corresponding video captions.\r\n\r\nImage source: [https://github.com/wenet-e2e/wenetspeech](https://github.com/wenet-e2e/wenetspeech)","description_withheld":null,"homepage":"https://wenet-e2e.github.io/WenetSpeech/","introduced_date":"2021-10-07","introduced_date_note":null,"introduced_by":{"paper":"/paper/wenetspeech-a-10000-hours-multi-domain","title":"WenetSpeech: A 10000+ Hours Multi-domain Mandarin Corpus for Speech Recognition","first_author":"BinBin Zhang","url":null},"license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"modalities":[{"name":"Speech","url":"/datasets/modality/speech"}],"tasks":[{"name":"Speech Recognition","url":"/task/speech-recognition","datasets_with_task":"/datasets/task/speech-recognition"}],"languages":[{"name":"Mandarin Chinese","url":"/datasets/language/mandarin-chinese"}],"variants":["WenetSpeech"],"data_loaders":[{"repo":"https://github.com/wenet-e2e/wenetspeech","url":"https://github.com/wenet-e2e/wenetspeech","frameworks":[]}],"num_papers_in_archive":58,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/speech-recognition-on-wenetspeech","task":"Speech Recognition","dataset_variant":"WenetSpeech","rows":8,"metrics":["Character Error Rate (CER)"],"first_row_in_archive_order":{"model":"Paraformer-large","paper":"/paper/funasr-a-fundamental-end-to-end-speech","metrics":{"Character Error Rate (CER)":"6.97"},"code_links":[{"title":"alibaba-damo-academy/FunASR","url":"https://github.com/alibaba-damo-academy/FunASR"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/zipformer-a-faster-and-better-encoder-for","title":"Zipformer: A faster and better encoder for automatic speech recognition","date":"2023-10-17","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/funasr-a-fundamental-end-to-end-speech","title":"FunASR: A Fundamental End-to-End Speech Recognition Toolkit","date":"2023-05-18","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/3m-multi-loss-multi-path-and-multi-level","title":"3M: Multi-loss, Multi-path and Multi-level Neural Networks for speech recognition","date":"2022-04-07","rows_on_this_dataset":3,"code_links":1,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":1,"samples_ran":1,"samples_unverified":0,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/wenetspeech-a-10000-hours-multi-domain","title":"WenetSpeech: A 10000+ Hours Multi-domain Mandarin Corpus for Speech Recognition","date":"2021-10-07","rows_on_this_dataset":3,"code_links":2,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":8,"samples_ran":8,"samples_unverified":0,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}}],"syntology_totals":{"read_at":"2026-09-25T09:33:49+00:00","papers_with_samples":2,"samples_harvested":9,"samples_ran":9,"samples_unverified":0,"pointer_only_for_licence":0,"papers_with_no_sample_that_ran":0,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}