{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/continuous-speech-separation-with-conformer","title":"Continuous Speech Separation with Conformer","arxiv_id":"2008.05773","date":"2020-08-13","proceeding":null,"authors":["Sanyuan Chen","Yu Wu","Zhuo Chen","Jian Wu","Jinyu Li","Takuya Yoshioka","Chengyi Wang","Shujie Liu","Ming Zhou"],"abstract":"Continuous speech separation plays a vital role in complicated speech related tasks such as conversation transcription. The separation model extracts a single speaker signal from a mixed speech. In this paper, we use transformer and conformer in lieu of recurrent neural networks in the separation system, as we believe capturing global information with the self-attention based method is crucial for the speech separation. Evaluating on the LibriCSS dataset, the conformer separation model achieves state of the art results, with a relative 23.5% word error rate (WER) reduction from bi-directional LSTM (BLSTM) in the utterance-wise evaluation and a 15.4% WER reduction in the continuous evaluation.","url_abs":"https://arxiv.org/abs/2008.05773v2","url_pdf":"https://arxiv.org/pdf/2008.05773v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"continuous-speech-separation-with-conformer","repo_url":"https://github.com/Sanyuan-Chen/CSS_with_Conformer","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"speech-separation","task_name":"Speech Separation"}],"methods":[{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/speech-separation-on-libricss","task":"Speech Separation","dataset":"LibriCSS","model":"Conformer (large)","rank_in_archive_order":1,"of":2,"metrics":{"0L":"5.0","0S":"5.4","10%":"7.5","20%":"10.7","30%":"13.8","40%":"17.1"},"uses_additional_data":true},{"leaderboard":"/sota/speech-separation-on-libricss","task":"Speech Separation","dataset":"LibriCSS","model":"Conformer (base)","rank_in_archive_order":2,"of":2,"metrics":{"0L":"5.4","0S":"5.6","10%":"8.2","20%":"11.8","30%":"15.5","40%":"18.9"},"uses_additional_data":true}],"syntology":{"syntology_url":"https://syntology.ai/paper/2008.05773","atlas_url":"https://app.syntology.ai/?focus=2008.05773","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}