{"url":"/dataset/aishell-2","name":"AISHELL-2","full_name":null,"description_markdown":"AISHELL-2 contains 1000 hours of clean read-speech data from iOS is free for academic usage. \r\n\r\nSource: [AISHELL-2: Transforming Mandarin ASR Research Into Industrial Scale](/paper/aishell-2-transforming-mandarin-asr-research)\r\nImage Source: [https://arxiv.org/pdf/1808.10583v2.pdf](https://arxiv.org/pdf/1808.10583v2.pdf)","description_withheld":null,"homepage":"https://github.com/kaldi-asr/kaldi/tree/master/egs/aishell2","introduced_date":null,"introduced_date_note":null,"introduced_by":{"paper":"/paper/aishell-2-transforming-mandarin-asr-research","title":"AISHELL-2: Transforming Mandarin ASR Research Into Industrial Scale","first_author":"Jiayu Du","url":null},"license":{"name":"Custom","url":"https://github.com/kaldi-asr/kaldi/tree/master/egs/aishell2"},"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":["AISHELL-2","AISHELL-2 Test IOS","AISHELL-2 Test Android","AISHELL-2 Test Mic"],"data_loaders":[{"repo":"https://github.com/kaldi-asr/kaldi","url":"https://github.com/kaldi-asr/kaldi","frameworks":[]}],"num_papers_in_archive":75,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/speech-recognition-on-aishell-2","task":"Speech Recognition","dataset_variant":"AISHELL-2","rows":2,"metrics":["Word Error Rate (WER)"],"first_row_in_archive_order":{"model":"Paraformer-large","paper":"/paper/funasr-a-fundamental-end-to-end-speech","metrics":{"Word Error Rate (WER)":"2.85"},"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"},{"leaderboard":"/sota/speech-recognition-on-aishell-2-test-android-1","task":"Speech Recognition","dataset_variant":"AISHELL-2 Test Android","rows":1,"metrics":["Word Error Rate (WER)"],"first_row_in_archive_order":{"model":"Qwen-Audio","paper":"/paper/qwen-audio-advancing-universal-audio","metrics":{"Word Error Rate (WER)":"3.3"},"code_links":[{"title":"alibaba-damo-academy/FunASR","url":"https://github.com/alibaba-damo-academy/FunASR"},{"title":"qwenlm/qwen-audio","url":"https://github.com/qwenlm/qwen-audio"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/speech-recognition-on-aishell-2-test-ios","task":"Speech Recognition","dataset_variant":"AISHELL-2 Test IOS","rows":1,"metrics":["Word Error Rate (WER)"],"first_row_in_archive_order":{"model":"Qwen-Audio","paper":"/paper/qwen-audio-advancing-universal-audio","metrics":{"Word Error Rate (WER)":"3.1"},"code_links":[{"title":"alibaba-damo-academy/FunASR","url":"https://github.com/alibaba-damo-academy/FunASR"},{"title":"qwenlm/qwen-audio","url":"https://github.com/qwenlm/qwen-audio"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/speech-recognition-on-aishell-2-test-mic-1","task":"Speech Recognition","dataset_variant":"AISHELL-2 Test Mic","rows":1,"metrics":["Word Error Rate (WER)"],"first_row_in_archive_order":{"model":"Qwen-Audio","paper":"/paper/qwen-audio-advancing-universal-audio","metrics":{"Word Error Rate (WER)":"3.3"},"code_links":[{"title":"alibaba-damo-academy/FunASR","url":"https://github.com/alibaba-damo-academy/FunASR"},{"title":"qwenlm/qwen-audio","url":"https://github.com/qwenlm/qwen-audio"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/qwen-audio-advancing-universal-audio","title":"Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models","date":"2023-11-14","rows_on_this_dataset":3,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":7,"samples_ran":5,"samples_unverified":2,"pointer_only_for_licence":7,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"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":2,"code_links":1,"syntology":null}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":1,"samples_harvested":7,"samples_ran":5,"samples_unverified":2,"pointer_only_for_licence":7,"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."}