{"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/unpaired-speech-enhancement-by-acoustic-and","title":"Unpaired Speech Enhancement by Acoustic and Adversarial Supervision for Speech Recognition","arxiv_id":"1811.02182","date":"2018-11-06","proceeding":null,"authors":["Geonmin Kim","Hwaran Lee","Bo-Kyeong Kim","Sang-Hoon Oh","Soo-Young Lee"],"abstract":"Many speech enhancement methods try to learn the relationship between noisy\nand clean speech, obtained using an acoustic room simulator. We point out\nseveral limitations of enhancement methods relying on clean speech targets; the\ngoal of this work is proposing an alternative learning algorithm, called\nacoustic and adversarial supervision (AAS). AAS makes the enhanced output both\nmaximizing the likelihood of transcription on the pre-trained acoustic model\nand having general characteristics of clean speech, which improve\ngeneralization on unseen noisy speeches. We employ the connectionist temporal\nclassification and the unpaired conditional boundary equilibrium generative\nadversarial network as the loss function of AAS. AAS is tested on two datasets\nincluding additive noise without and with reverberation, Librispeech + DEMAND\nand CHiME-4. By visualizing the enhanced speech with different loss\ncombinations, we demonstrate the role of each supervision. AAS achieves a lower\nword error rate than other state-of-the-art methods using the clean speech\ntarget in both datasets.","url_abs":"http://arxiv.org/abs/1811.02182v1","url_pdf":"http://arxiv.org/pdf/1811.02182v1.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":"unpaired-speech-enhancement-by-acoustic-and","repo_url":"https://github.com/gmkim90/AAS_enhancement","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":null,"task_name":"Generative Adversarial Network"},{"task_slug":"speech-enhancement","task_name":"Speech Enhancement"},{"task_slug":"speech-recognition","task_name":"Speech Recognition"},{"task_slug":"speech-recognition-1","task_name":"speech-recognition"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}