{"id":"https://openalex.org/W3097224945","doi":"https://doi.org/10.18653/v1/2021.naacl-main.154","title":"Align-Refine: Non-Autoregressive Speech Recognition via Iterative Realignment","display_name":"Align-Refine: Non-Autoregressive Speech Recognition via Iterative Realignment","publication_year":2021,"publication_date":"2021-01-01","ids":{"openalex":"https://openalex.org/W3097224945","doi":"https://doi.org/10.18653/v1/2021.naacl-main.154","mag":"3097224945"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2021.naacl-main.154","doi":"https://doi.org/10.18653/v1/2021.naacl-main.154","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2021.naacl-main.154","pdf_url":"https://aclanthology.org/2021.naacl-main.154.pdf","source":{"id":"https://openalex.org/S4363608774","display_name":"Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://aclanthology.org/2021.naacl-main.154.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5112009475","display_name":"Ethan A. Chi","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ethan A. Chi","raw_affiliation_strings":["Stanford University ()"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Stanford University ()","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037283647","display_name":"Juli\u00e1n Salazar","orcid":"https://orcid.org/0000-0003-0943-7594","observed_orcids":["https://orcid.org/0000-0003-0943-7594"]},"institutions":[{"id":"https://openalex.org/I1311688040","display_name":"Amazon (United States)","ror":"https://ror.org/04mv4n011","country_code":"US","type":"company","lineage":["https://openalex.org/I1311688040"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Julian Salazar","raw_affiliation_strings":["Amazon"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Amazon","institution_ids":["https://openalex.org/I1311688040"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5050325468","display_name":"Katrin Kirchhoff","orcid":"https://orcid.org/0000-0002-6645-6030","observed_orcids":["https://orcid.org/0000-0002-6645-6030"]},"institutions":[{"id":"https://openalex.org/I1311688040","display_name":"Amazon (United States)","ror":"https://ror.org/04mv4n011","country_code":"US","type":"company","lineage":["https://openalex.org/I1311688040"]},{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Katrin Kirchhoff","raw_affiliation_strings":["Stanford University","Amazon"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Stanford University","institution_ids":["https://openalex.org/I97018004"]},{"raw_affiliation_string":"Amazon","institution_ids":["https://openalex.org/I1311688040"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":3.4572,"has_fulltext":true,"cited_by_count":37,"citation_normalized_percentile":{"value":0.94143086,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":92,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"1920","last_page":"1927"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":9.999999747378752e-05,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/latent-alignment","display_name":"latent alignment","score":0.6840000152587891},{"id":"https://openalex.org/keywords/iterative-decoding","display_name":"iterative decoding","score":0.781000018119812},{"id":"https://openalex.org/keywords/decoding-speedup","display_name":"decoding speedup","score":0.7250000238418579},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"speech recognition","score":0.3230000138282776}],"concepts":[{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.8947515487670898},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7385957837104797},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.6504154205322266},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5890948176383972},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.5674816370010376},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.5627756714820862},{"id":"https://openalex.org/C8521452","wikidata":"https://www.wikidata.org/wiki/Q203790","display_name":"Connectionism","level":3,"score":0.5526825785636902},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4655148983001709},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.44313159584999084},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.41928544640541077},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.35348770022392273},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.22298404574394226},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.12810805439949036},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.08866140246391296},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.0706823468208313},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.0},{"id":"https://openalex.org/C54355233","wikidata":"https://www.wikidata.org/wiki/Q7162","display_name":"Genetics","level":1,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C119599485","wikidata":"https://www.wikidata.org/wiki/Q43035","display_name":"Electrical engineering","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.18653/v1/2021.naacl-main.154","doi":"https://doi.org/10.18653/v1/2021.naacl-main.154","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2021.naacl-main.154","pdf_url":"https://aclanthology.org/2021.naacl-main.154.pdf","source":{"id":"https://openalex.org/S4363608774","display_name":"Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies","raw_type":"proceedings-article"},{"id":"mag:3097224945","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/pdf/2010.14233.pdf","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null}],"best_oa_location":{"id":"doi:10.18653/v1/2021.naacl-main.154","doi":"https://doi.org/10.18653/v1/2021.naacl-main.154","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2021.naacl-main.154","pdf_url":"https://aclanthology.org/2021.naacl-main.154.pdf","source":{"id":"https://openalex.org/S4363608774","display_name":"Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies","raw_type":"proceedings-article"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W3097224945.pdf","grobid_xml":"https://content.openalex.org/works/W3097224945.grobid-xml"},"referenced_works_count":30,"referenced_works":["https://openalex.org/W1494198834","https://openalex.org/W2024490156","https://openalex.org/W2127141656","https://openalex.org/W2327501763","https://openalex.org/W2404126548","https://openalex.org/W2520160253","https://openalex.org/W2626778328","https://openalex.org/W2767206889","https://openalex.org/W2889326796","https://openalex.org/W2892009249","https://openalex.org/W2896457183","https://openalex.org/W2911291251","https://openalex.org/W2936774411","https://openalex.org/W2937808806","https://openalex.org/W2962780374","https://openalex.org/W2963341956","https://openalex.org/W2963434219","https://openalex.org/W2964308564","https://openalex.org/W2972818416","https://openalex.org/W2979636403","https://openalex.org/W3007746272","https://openalex.org/W3015537910","https://openalex.org/W3024308166","https://openalex.org/W3035812575","https://openalex.org/W3095173472","https://openalex.org/W3097438275","https://openalex.org/W3097882114","https://openalex.org/W3112157188","https://openalex.org/W3162919436","https://openalex.org/W4287628715"],"related_works":["https://openalex.org/W4223950106","https://openalex.org/W7166739935","https://openalex.org/W4224931714","https://openalex.org/W3095197298","https://openalex.org/W7131908762","https://openalex.org/W4361990931","https://openalex.org/W4303020902","https://openalex.org/W7165755494","https://openalex.org/W4372266908","https://openalex.org/W3198068643"],"abstract_inverted_index":{"Non-autoregressive":[0],"encoder-decoder":[1],"models":[2,20],"greatly":[3],"improve":[4],"decoding":[5,19,82],"speed":[6],"over":[7],"autoregressive":[8,77],"models,":[9],"at":[10,32,106],"the":[11,25,33,71],"expense":[12],"of":[13,27,35,92],"generation":[14],"quality.To":[15],"mitigate":[16],"this,":[17],"iterative":[18,43],"repeatedly":[21],"infill":[22],"or":[23],"refine":[24],"proposal":[26],"a":[28,80],"non-autoregressive":[29],"model.However,":[30],"editing":[31],"level":[34],"output":[36],"sequences":[37],"limits":[38],"model":[39],"flexibility.We":[40],"instead":[41],"propose":[42],"realignment,":[44],"which":[45,63],"by":[46],"refining":[47],"latent":[48],"alignments":[49],"allows":[50],"more":[51],"flexible":[52],"edits":[53],"in":[54,100],"fewer":[55],"steps.Our":[56],"model,":[57],"Align-Refine,":[58],"is":[59],"an":[60,76,88],"end-to-end":[61],"Transformer":[62],"iteratively":[64],"realigns":[65],"connectionist":[66],"temporal":[67],"classification":[68],"(CTC)":[69],"alignments.On":[70],"WSJ":[72],"dataset,":[73],"Align-Refine":[74],"matches":[75],"baseline":[78],"with":[79],"14\u00d7":[81],"speedup;":[83],"on":[84,97],"LibriSpeech,":[85],"we":[86],"reach":[87],"LM-free":[89],"testother":[90],"WER":[91],"9.0%":[93],"(19%":[94],"relative":[95],"improvement":[96],"comparable":[98],"work)":[99],"three":[101],"iterations.We":[102],"release":[103],"our":[104],"code":[105],"https://github.com/":[107],"amazon-research/align-refine.":[108]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":5},{"year":2024,"cited_by_count":2},{"year":2023,"cited_by_count":9},{"year":2022,"cited_by_count":10},{"year":2021,"cited_by_count":10}],"updated_date":"2026-11-03T10:50:19.383937","created_date":"2020-11-09T00:00:00"}