{"id":"https://openalex.org/W3163907627","doi":"https://doi.org/10.1109/icassp39728.2021.9413803","title":"FastEmit: Low-Latency Streaming ASR with Sequence-Level Emission Regularization","display_name":"FastEmit: Low-Latency Streaming ASR with Sequence-Level Emission Regularization","publication_year":2021,"publication_date":"2021-05-13","ids":{"openalex":"https://openalex.org/W3163907627","doi":"https://doi.org/10.1109/icassp39728.2021.9413803","mag":"3163907627"},"language":"en","primary_location":{"id":"doi:10.1109/icassp39728.2021.9413803","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp39728.2021.9413803","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5001205723","display_name":"Jiahui Yu","orcid":"https://orcid.org/0000-0003-1314-2481"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Jiahui Yu","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5027763497","display_name":"Chung\u2010Cheng Chiu","orcid":"https://orcid.org/0000-0001-9729-4778"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Chung-Cheng Chiu","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100374448","display_name":"Bo Li","orcid":"https://orcid.org/0000-0002-6711-3603"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Bo Li","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5001306222","display_name":"Shuo-Yiin Chang","orcid":null},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shuo-yiin Chang","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5070513394","display_name":"Tara N. Sainath","orcid":"https://orcid.org/0000-0002-4126-6556"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Tara N. Sainath","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101319167","display_name":"Yanzhang He","orcid":null},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yanzhang He","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5000078382","display_name":"Arun Narayanan","orcid":"https://orcid.org/0009-0008-3325-8928"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Arun Narayanan","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100750907","display_name":"Wei Han","orcid":"https://orcid.org/0000-0002-4201-9645"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Wei Han","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5070108948","display_name":"Anmol Gulati","orcid":"https://orcid.org/0009-0007-2077-9583"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Anmol Gulati","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5010253402","display_name":"Yonghui Wu","orcid":"https://orcid.org/0000-0002-6780-6135"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yonghui Wu","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5112901893","display_name":"Ruoming Pang","orcid":null},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ruoming Pang","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":11,"corresponding_author_ids":["https://openalex.org/A5001205723"],"corresponding_institution_ids":["https://openalex.org/I1291425158"],"apc_list":null,"apc_paid":null,"fwci":7.9774,"has_fulltext":false,"cited_by_count":64,"citation_normalized_percentile":{"value":0.97864022,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":89,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"6004","last_page":"6008"},"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":0.9993000030517578,"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/T11309","display_name":"Music and Audio Processing","score":0.9980000257492065,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.748205304145813},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.6287827491760254},{"id":"https://openalex.org/keywords/transducer","display_name":"Transducer","score":0.6214471459388733},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.5612742304801941},{"id":"https://openalex.org/keywords/regularization","display_name":"Regularization (linguistics)","score":0.5311918258666992},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.503600537776947},{"id":"https://openalex.org/keywords/word-error-rate","display_name":"Word error rate","score":0.48945653438568115},{"id":"https://openalex.org/keywords/real-time-computing","display_name":"Real-time computing","score":0.3640841841697693},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.31076711416244507},{"id":"https://openalex.org/keywords/acoustics","display_name":"Acoustics","score":0.16137945652008057},{"id":"https://openalex.org/keywords/telecommunications","display_name":"Telecommunications","score":0.12090447545051575},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.10190394520759583}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.748205304145813},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.6287827491760254},{"id":"https://openalex.org/C56318395","wikidata":"https://www.wikidata.org/wiki/Q215928","display_name":"Transducer","level":2,"score":0.6214471459388733},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.5612742304801941},{"id":"https://openalex.org/C2776135515","wikidata":"https://www.wikidata.org/wiki/Q17143721","display_name":"Regularization (linguistics)","level":2,"score":0.5311918258666992},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.503600537776947},{"id":"https://openalex.org/C40969351","wikidata":"https://www.wikidata.org/wiki/Q3516228","display_name":"Word error rate","level":2,"score":0.48945653438568115},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.3640841841697693},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.31076711416244507},{"id":"https://openalex.org/C24890656","wikidata":"https://www.wikidata.org/wiki/Q82811","display_name":"Acoustics","level":1,"score":0.16137945652008057},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.12090447545051575},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.10190394520759583},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp39728.2021.9413803","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp39728.2021.9413803","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/16","display_name":"Peace, Justice and strong institutions","score":0.5699999928474426}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":33,"referenced_works":["https://openalex.org/W1494198834","https://openalex.org/W1828163288","https://openalex.org/W1855892484","https://openalex.org/W2742061524","https://openalex.org/W2892209435","https://openalex.org/W2928941594","https://openalex.org/W2936774411","https://openalex.org/W2962760690","https://openalex.org/W2962893195","https://openalex.org/W2963122170","https://openalex.org/W2963747784","https://openalex.org/W2963920996","https://openalex.org/W2973122799","https://openalex.org/W3008898571","https://openalex.org/W3015194534","https://openalex.org/W3015927303","https://openalex.org/W3016010032","https://openalex.org/W3021469861","https://openalex.org/W3021515889","https://openalex.org/W3025165719","https://openalex.org/W3092970820","https://openalex.org/W3093346109","https://openalex.org/W3095173472","https://openalex.org/W3095994005","https://openalex.org/W3096524176","https://openalex.org/W3096702180","https://openalex.org/W3097777922","https://openalex.org/W4288088457","https://openalex.org/W6638749077","https://openalex.org/W6639156005","https://openalex.org/W6760633627","https://openalex.org/W6769806307","https://openalex.org/W6784800133"],"related_works":["https://openalex.org/W4388335561","https://openalex.org/W2970530566","https://openalex.org/W2967478618","https://openalex.org/W2997152889","https://openalex.org/W4385572700","https://openalex.org/W4307309205","https://openalex.org/W4288261899","https://openalex.org/W4385009901","https://openalex.org/W4384820447","https://openalex.org/W4389115113"],"abstract_inverted_index":{"Streaming":[0],"automatic":[1],"speech":[2],"recognition":[3],"(ASR)":[4],"aims":[5],"to":[6,122,181,191],"emit":[7],"each":[8],"hypothesized":[9],"word":[10,26,78],"as":[11,15,23],"quickly":[12],"and":[13,37,41,74,109,150],"accurately":[14],"possible.":[16],"However,":[17],"emitting":[18],"fast":[19],"without":[20],"degrading":[21],"quality,":[22],"measured":[24],"by":[25,49,133],"error":[27],"rate":[28],"(WER),":[29],"is":[30,119],"highly":[31],"challenging.":[32],"Existing":[33],"approaches":[34,68],"including":[35,142],"Early":[36],"Late":[38],"Penalties":[39],"[1]":[40],"Constrained":[42],"Alignments":[43],"[2],":[44],"[3]":[45],"penalize":[46],"emission":[47,92],"delay":[48],"manipulating":[50],"per-token":[51],"or":[52],"per-frame":[53],"probability":[54,104],"prediction":[55],"in":[56,64,105],"sequence":[57],"transducer":[58,107,127],"models":[59,128],"[4].":[60],"While":[61],"being":[62],"successful":[63],"reducing":[65],"delay,":[66],"these":[67],"suffer":[69],"from":[70,81,179,189],"significant":[71],"accuracy":[72,163,178],"regression":[73],"also":[75,174],"require":[76,112],"additional":[77],"alignment":[79],"information":[80],"an":[82],"existing":[83],"model.":[84],"In":[85],"this":[86],"work,":[87],"we":[88],"propose":[89],"a":[90,168],"sequence-level":[91,124],"regularization":[93,100],"method,":[94],"named":[95],"FastEmit,":[96],"that":[97,117],"applies":[98],"latency":[99,158,188],"directly":[101],"on":[102,136,167,194],"per-sequence":[103],"training":[106],"models,":[108],"does":[110],"not":[111],"any":[113],"alignment.":[114],"We":[115,153],"demonstrate":[116],"FastEmit":[118,173],"more":[120],"suitable":[121],"the":[123],"optimization":[125],"of":[126],"[4]":[129],"for":[130],"streaming":[131,139,176],"ASR":[132,140,177],"applying":[134],"it":[135],"various":[137],"end-to-end":[138],"networks":[141],"RNN-Transducer":[143],"[5],":[144],"Transformer-Transducer":[145],"[6],":[146],"[7],":[147],"ConvNet-Transducer":[148],"[8]":[149],"Conformer-Transducer":[151],"[9].":[152],"achieve":[154],"150":[155],"~":[156],"300ms":[157],"reduction":[159],"with":[160],"significantly":[161],"better":[162],"over":[164],"previous":[165],"techniques":[166],"Voice":[169],"Search":[170],"test":[171],"set.":[172],"improves":[175],"4.4%/8.9%":[180],"3.1%/7.5%":[182],"WER,":[183],"meanwhile":[184],"reduces":[185],"90th":[186],"percentile":[187],"210ms":[190],"only":[192],"30ms":[193],"LibriSpeech.":[195]},"counts_by_year":[{"year":2025,"cited_by_count":6},{"year":2024,"cited_by_count":7},{"year":2023,"cited_by_count":27},{"year":2022,"cited_by_count":12},{"year":2021,"cited_by_count":11},{"year":2020,"cited_by_count":1}],"updated_date":"2026-04-04T06:10:10.580331","created_date":"2025-10-10T00:00:00"}