{"id":"https://openalex.org/W2773723600","doi":"https://doi.org/10.1109/icassp.2018.8461809","title":"Minimum Word Error Rate Training for Attention-Based Sequence-to-Sequence Models","display_name":"Minimum Word Error Rate Training for Attention-Based Sequence-to-Sequence Models","publication_year":2018,"publication_date":"2018-04-01","ids":{"openalex":"https://openalex.org/W2773723600","doi":"https://doi.org/10.1109/icassp.2018.8461809","mag":"2773723600"},"language":"en","primary_location":{"id":"doi:10.1109/icassp.2018.8461809","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp.2018.8461809","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/1712.01818","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5032640894","display_name":"Rohit Prabhavalkar","orcid":"https://orcid.org/0000-0001-5331-6058"},"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":"Rohit Prabhavalkar","raw_affiliation_strings":["Google Inc","GOOGLE INC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google Inc","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"GOOGLE INC","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 Inc","GOOGLE INC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google Inc","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"GOOGLE INC","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 Inc","GOOGLE INC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google Inc","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"GOOGLE INC","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5111455517","display_name":"Patrick Nguyen","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":"Patrick Nguyen","raw_affiliation_strings":["Google Inc","GOOGLE INC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google Inc","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"GOOGLE INC","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100715424","display_name":"Zhifeng Chen","orcid":"https://orcid.org/0000-0001-8631-2424"},"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":"Zhifeng Chen","raw_affiliation_strings":["Google Inc","GOOGLE INC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google Inc","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"GOOGLE INC","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 Inc","GOOGLE INC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google Inc","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"GOOGLE INC","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5023098355","display_name":"Anjuli Kannan","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":"Anjuli Kannan","raw_affiliation_strings":["Google Inc","GOOGLE INC"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google Inc","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"GOOGLE INC","institution_ids":["https://openalex.org/I1291425158"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I1291425158"],"apc_list":null,"apc_paid":null,"fwci":0.5659,"has_fulltext":true,"cited_by_count":5,"citation_normalized_percentile":{"value":0.69750934,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"4839","last_page":"4843"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9998999834060669,"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":0.9998999834060669,"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/T10028","display_name":"Topic Modeling","score":0.9987999796867371,"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/T11309","display_name":"Music and Audio Processing","score":0.9966999888420105,"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/word-error-rate","display_name":"Word error rate","score":0.7739983797073364},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7313453555107117},{"id":"https://openalex.org/keywords/discriminative-model","display_name":"Discriminative model","score":0.7036285996437073},{"id":"https://openalex.org/keywords/word","display_name":"Word (group theory)","score":0.6194027066230774},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5946581363677979},{"id":"https://openalex.org/keywords/sequence","display_name":"Sequence (biology)","score":0.5833064913749695},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5012619495391846},{"id":"https://openalex.org/keywords/bayes-theorem","display_name":"Bayes' theorem","score":0.4472386837005615},{"id":"https://openalex.org/keywords/principle-of-maximum-entropy","display_name":"Principle of maximum entropy","score":0.44606462121009827},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.43180084228515625},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.36327338218688965},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.3510802984237671},{"id":"https://openalex.org/keywords/bayesian-probability","display_name":"Bayesian probability","score":0.22651103138923645},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.1332549750804901}],"concepts":[{"id":"https://openalex.org/C40969351","wikidata":"https://www.wikidata.org/wiki/Q3516228","display_name":"Word error rate","level":2,"score":0.7739983797073364},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7313453555107117},{"id":"https://openalex.org/C97931131","wikidata":"https://www.wikidata.org/wiki/Q5282087","display_name":"Discriminative model","level":2,"score":0.7036285996437073},{"id":"https://openalex.org/C90805587","wikidata":"https://www.wikidata.org/wiki/Q10944557","display_name":"Word (group theory)","level":2,"score":0.6194027066230774},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5946581363677979},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.5833064913749695},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5012619495391846},{"id":"https://openalex.org/C207201462","wikidata":"https://www.wikidata.org/wiki/Q182505","display_name":"Bayes' theorem","level":3,"score":0.4472386837005615},{"id":"https://openalex.org/C9679016","wikidata":"https://www.wikidata.org/wiki/Q1417473","display_name":"Principle of maximum entropy","level":2,"score":0.44606462121009827},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.43180084228515625},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.36327338218688965},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3510802984237671},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.22651103138923645},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.1332549750804901},{"id":"https://openalex.org/C187736073","wikidata":"https://www.wikidata.org/wiki/Q2920921","display_name":"Management","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/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0},{"id":"https://openalex.org/C54355233","wikidata":"https://www.wikidata.org/wiki/Q7162","display_name":"Genetics","level":1,"score":0.0}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.1109/icassp.2018.8461809","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp.2018.8461809","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:1712.01818","is_oa":true,"landing_page_url":"http://arxiv.org/abs/1712.01818","pdf_url":"https://arxiv.org/pdf/1712.01818","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,"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":null,"raw_type":"text"},{"id":"mag:2773723600","is_oa":true,"landing_page_url":"https://arxiv.org/pdf/1712.01818.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,"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},{"id":"doi:10.48550/arxiv.1712.01818","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.1712.01818","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,"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":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:1712.01818","is_oa":true,"landing_page_url":"http://arxiv.org/abs/1712.01818","pdf_url":"https://arxiv.org/pdf/1712.01818","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,"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":null,"raw_type":"text"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/10","score":0.7200000286102295,"display_name":"Reduced inequalities"}],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W2773723600.pdf","grobid_xml":"https://content.openalex.org/works/W2773723600.grobid-xml"},"referenced_works_count":52,"referenced_works":["https://openalex.org/W1828163288","https://openalex.org/W1987238397","https://openalex.org/W2064675550","https://openalex.org/W2101105183","https://openalex.org/W2102113734","https://openalex.org/W2119717200","https://openalex.org/W2121879602","https://openalex.org/W2127141656","https://openalex.org/W2131342762","https://openalex.org/W2131774270","https://openalex.org/W2271840356","https://openalex.org/W2291513470","https://openalex.org/W2327501763","https://openalex.org/W2515439472","https://openalex.org/W2545177271","https://openalex.org/W2617258110","https://openalex.org/W2622301280","https://openalex.org/W2626778328","https://openalex.org/W2746192915","https://openalex.org/W2749514303","https://openalex.org/W2750499125","https://openalex.org/W2809161641","https://openalex.org/W2949888546","https://openalex.org/W2950304420","https://openalex.org/W2951781666","https://openalex.org/W2962765220","https://openalex.org/W2962826786","https://openalex.org/W2963248296","https://openalex.org/W2963414781","https://openalex.org/W2963920996","https://openalex.org/W2964121744","https://openalex.org/W2964308564","https://openalex.org/W6621543089","https://openalex.org/W6629052376","https://openalex.org/W6631190155","https://openalex.org/W6638749077","https://openalex.org/W6647191204","https://openalex.org/W6675365184","https://openalex.org/W6676562027","https://openalex.org/W6676981579","https://openalex.org/W6678555148","https://openalex.org/W6679434410","https://openalex.org/W6679436768","https://openalex.org/W6680402377","https://openalex.org/W6685322675","https://openalex.org/W6702097088","https://openalex.org/W6727193155","https://openalex.org/W6739901393","https://openalex.org/W6743440867","https://openalex.org/W6746892368","https://openalex.org/W6747398299","https://openalex.org/W6898505805"],"related_works":["https://openalex.org/W2963747784","https://openalex.org/W2962765220","https://openalex.org/W2133564850","https://openalex.org/W1959963859","https://openalex.org/W2951444698","https://openalex.org/W1528470941","https://openalex.org/W2044684117","https://openalex.org/W2148386535","https://openalex.org/W2972698273","https://openalex.org/W2158339352","https://openalex.org/W2963727906","https://openalex.org/W3135418384","https://openalex.org/W2151129782","https://openalex.org/W2093478938","https://openalex.org/W2162432783","https://openalex.org/W2156427312","https://openalex.org/W2104036986","https://openalex.org/W2975068042","https://openalex.org/W1878324503","https://openalex.org/W2541906951"],"abstract_inverted_index":{"Sequence-to-sequence":[0],"models,":[1],"such":[2,54],"as":[3,55],"attention-based":[4,78,153],"models":[5,79,154],"in":[6,34],"automatic":[7],"speech":[8],"recognition":[9],"(ASR),":[10],"are":[11,63],"typically":[12],"trained":[13],"to":[14,21,67,76,80,117,139,142,149],"optimize":[15],"the":[16,23,26,56,70,94,104,122,131,143,157],"cross-entropy":[17],"criterion":[18],"which":[19,51,62,92,114,155],"corresponds":[20],"improving":[22],"log-likelihood":[24],"of":[25,36,97,111,159],"data.":[27],"However,":[28],"system":[29,165],"performance":[30,136,158],"is":[31],"usually":[32],"measured":[33],"terms":[35],"word":[37,84,98],"error":[38,85],"rate":[39],"(WER),":[40],"not":[41],"log-likelihood.":[42],"Traditional":[43],"ASR":[44],"systems":[45],"benefit":[46],"from":[47,103],"discriminative":[48,163],"sequence":[49],"training":[50,133],"optimizes":[52],"criteria":[53],"state-level":[57],"minimum":[58],"Bayes":[59],"risk":[60],"(sMBR)":[61],"more":[64,119],"closely":[65],"related":[66],"WER.":[68],"In":[69,125],"present":[71],"work,":[72],"we":[73,115,128],"explore":[74],"techniques":[75],"train":[77,150],"directly":[81],"minimize":[82],"expected":[83,95],"rate.":[86],"We":[87],"consider":[88],"two":[89],"loss":[90],"functions":[91],"approximate":[93],"number":[96],"errors:":[99],"either":[100],"by":[101,107,137],"sampling":[102],"model,":[105],"or":[106],"using":[108],"N-best":[109],"lists":[110],"decoded":[112],"hypotheses,":[113],"find":[116,129],"be":[118],"effective":[120],"than":[121],"sampling-based":[123],"method.":[124],"experimental":[126],"evaluations,":[127],"that":[130],"proposed":[132],"procedure":[134],"improves":[135],"up":[138],"8.2%":[140],"relative":[141],"baseline":[144],"system.":[145],"This":[146],"allows":[147],"us":[148],"grapheme-based,":[151],"uni-directional":[152],"match":[156],"a":[160,167],"traditional,":[161],"state-of-the-art,":[162],"sequence-trained":[164],"on":[166],"mobile":[168],"voice-search":[169],"task.":[170]},"counts_by_year":[{"year":2021,"cited_by_count":1},{"year":2019,"cited_by_count":1},{"year":2018,"cited_by_count":3}],"updated_date":"2026-08-05T07:39:15.569665","created_date":"2025-10-10T00:00:00"}