{"id":"https://openalex.org/W4221160683","doi":"https://doi.org/10.1109/icassp43922.2022.9747118","title":"RescoreBERT: Discriminative Speech Recognition Rescoring With Bert","display_name":"RescoreBERT: Discriminative Speech Recognition Rescoring With Bert","publication_year":2022,"publication_date":"2022-04-27","ids":{"openalex":"https://openalex.org/W4221160683","doi":"https://doi.org/10.1109/icassp43922.2022.9747118"},"language":"en","primary_location":{"id":"doi:10.1109/icassp43922.2022.9747118","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp43922.2022.9747118","pdf_url":null,"source":{"id":"https://openalex.org/S4363607702","display_name":"ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},"type":"article","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2202.01094","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101528687","display_name":"Liyan Xu","orcid":"https://orcid.org/0000-0001-6531-1123"},"institutions":[{"id":"https://openalex.org/I150468666","display_name":"Emory University","ror":"https://ror.org/03czfpz43","country_code":"US","type":"education","lineage":["https://openalex.org/I150468666"]},{"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":true,"raw_author_name":"Liyan Xu","raw_affiliation_strings":["Amazon Alexa AI,USA","Emory University, USA","Amazon Alexa AI, USA"],"affiliations":[{"raw_affiliation_string":"Amazon Alexa AI,USA","institution_ids":["https://openalex.org/I1311688040"]},{"raw_affiliation_string":"Emory University, USA","institution_ids":["https://openalex.org/I150468666"]},{"raw_affiliation_string":"Amazon Alexa AI, USA","institution_ids":["https://openalex.org/I1311688040"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5022646825","display_name":"Yile Gu","orcid":"https://orcid.org/0000-0003-3662-5056"},"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":"Yile Gu","raw_affiliation_strings":["Amazon Alexa AI,USA","Amazon Alexa AI, USA"],"affiliations":[{"raw_affiliation_string":"Amazon Alexa AI,USA","institution_ids":["https://openalex.org/I1311688040"]},{"raw_affiliation_string":"Amazon Alexa AI, USA","institution_ids":["https://openalex.org/I1311688040"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5052467988","display_name":"Jari Kolehmainen","orcid":"https://orcid.org/0000-0002-4174-6277"},"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":"Jari Kolehmainen","raw_affiliation_strings":["Amazon Alexa AI,USA","Amazon Alexa AI, USA"],"affiliations":[{"raw_affiliation_string":"Amazon Alexa AI,USA","institution_ids":["https://openalex.org/I1311688040"]},{"raw_affiliation_string":"Amazon Alexa AI, USA","institution_ids":["https://openalex.org/I1311688040"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083382561","display_name":"Haidar Khan","orcid":"https://orcid.org/0000-0003-1617-3208"},"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":"Haidar Khan","raw_affiliation_strings":["Amazon Alexa AI,USA","Amazon Alexa AI, USA"],"affiliations":[{"raw_affiliation_string":"Amazon Alexa AI,USA","institution_ids":["https://openalex.org/I1311688040"]},{"raw_affiliation_string":"Amazon Alexa AI, USA","institution_ids":["https://openalex.org/I1311688040"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5014054572","display_name":"Ankur Gandhe","orcid":null},"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":"Ankur Gandhe","raw_affiliation_strings":["Amazon Alexa AI,USA","Amazon Alexa AI, USA"],"affiliations":[{"raw_affiliation_string":"Amazon Alexa AI,USA","institution_ids":["https://openalex.org/I1311688040"]},{"raw_affiliation_string":"Amazon Alexa AI, USA","institution_ids":["https://openalex.org/I1311688040"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5110230355","display_name":"Ariya Rastrow","orcid":null},"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":"Ariya Rastrow","raw_affiliation_strings":["Amazon Alexa AI,USA","Amazon Alexa AI, USA"],"affiliations":[{"raw_affiliation_string":"Amazon Alexa AI,USA","institution_ids":["https://openalex.org/I1311688040"]},{"raw_affiliation_string":"Amazon Alexa AI, USA","institution_ids":["https://openalex.org/I1311688040"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5060979948","display_name":"Andreas Stolcke","orcid":"https://orcid.org/0000-0002-9925-905X"},"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":"Andreas Stolcke","raw_affiliation_strings":["Amazon Alexa AI,USA","Amazon Alexa AI, USA"],"affiliations":[{"raw_affiliation_string":"Amazon Alexa AI,USA","institution_ids":["https://openalex.org/I1311688040"]},{"raw_affiliation_string":"Amazon Alexa AI, USA","institution_ids":["https://openalex.org/I1311688040"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5109934165","display_name":"Ivan Bulyko","orcid":null},"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":"Ivan Bulyko","raw_affiliation_strings":["Amazon Alexa AI,USA","Amazon Alexa AI, USA"],"affiliations":[{"raw_affiliation_string":"Amazon Alexa AI,USA","institution_ids":["https://openalex.org/I1311688040"]},{"raw_affiliation_string":"Amazon Alexa AI, USA","institution_ids":["https://openalex.org/I1311688040"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":8,"corresponding_author_ids":["https://openalex.org/A5101528687"],"corresponding_institution_ids":["https://openalex.org/I1311688040","https://openalex.org/I150468666"],"apc_list":null,"apc_paid":null,"fwci":3.7797,"has_fulltext":false,"cited_by_count":36,"citation_normalized_percentile":{"value":0.94551622,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":97,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"6117","last_page":"6121"},"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/T11309","display_name":"Music and Audio Processing","score":0.9991999864578247,"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":0.9990000128746033,"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/computer-science","display_name":"Computer science","score":0.7695057392120361},{"id":"https://openalex.org/keywords/discriminative-model","display_name":"Discriminative model","score":0.7650998830795288},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.7251951694488525},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.36640554666519165}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7695057392120361},{"id":"https://openalex.org/C97931131","wikidata":"https://www.wikidata.org/wiki/Q5282087","display_name":"Discriminative model","level":2,"score":0.7650998830795288},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.7251951694488525},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.36640554666519165}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/icassp43922.2022.9747118","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp43922.2022.9747118","pdf_url":null,"source":{"id":"https://openalex.org/S4363607702","display_name":"ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2202.01094","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2202.01094","pdf_url":"https://arxiv.org/pdf/2202.01094","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2202.01094","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2202.01094","pdf_url":"https://arxiv.org/pdf/2202.01094","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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","display_name":"Reduced inequalities","score":0.7400000095367432}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":25,"referenced_works":["https://openalex.org/W1494198834","https://openalex.org/W1821462560","https://openalex.org/W1915251500","https://openalex.org/W2327501763","https://openalex.org/W2396464458","https://openalex.org/W2747917286","https://openalex.org/W2752047430","https://openalex.org/W2896457183","https://openalex.org/W2962760690","https://openalex.org/W2963174344","https://openalex.org/W2963747784","https://openalex.org/W2972320704","https://openalex.org/W2973122799","https://openalex.org/W2990391581","https://openalex.org/W3011339933","https://openalex.org/W3016188195","https://openalex.org/W3024308166","https://openalex.org/W3034775979","https://openalex.org/W3141464856","https://openalex.org/W3141961557","https://openalex.org/W6638523607","https://openalex.org/W6640059789","https://openalex.org/W6743726175","https://openalex.org/W6755207826","https://openalex.org/W6770528390"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2965546495","https://openalex.org/W4389116644","https://openalex.org/W2153315159","https://openalex.org/W3103844505","https://openalex.org/W259157601","https://openalex.org/W4205463238","https://openalex.org/W2761785940"],"abstract_inverted_index":{"Second-pass":[0],"rescoring":[1,27,58,90,196],"is":[2,13],"an":[3,138,172,194],"important":[4],"component":[5],"in":[6,44],"automatic":[7],"speech":[8],"recognition":[9],"(ASR)":[10],"systems":[11],"that":[12,118,181],"used":[14],"to":[15,86,95,127,190],"improve":[16],"the":[17,97,120,123,154],"outputs":[18],"from":[19,131,175],"a":[20,25,34,57,64,70,88,100,115,132,160,176],"first-pass":[21],"decoder":[22],"by":[23,150],"implementing":[24],"lattice":[26],"or":[28],"n-best":[29],"re-ranking.":[30],"While":[31],"pretraining":[32],"with":[33,92],"masked":[35],"language":[36,47],"model":[37,59,66,91],"(MLM)":[38],"objective":[39,72],"has":[40,52,78],"received":[41],"great":[42],"success":[43],"various":[45],"natural":[46],"understanding":[48],"(NLU)":[49],"tasks,":[50],"it":[51,182],"not":[53,79],"gained":[54],"traction":[55],"as":[56,74],"for":[60,110],"ASR.":[61,111],"Specifically,":[62,112],"training":[63,125],"bidirectional":[65,107],"like":[67],"BERT":[68,161],"on":[69,153,171],"discriminative":[71,101,124,140,164],"such":[73],"minimum":[75],"WER":[76,149,187],"(MWER)":[77],"been":[80],"explored.":[81],"Here":[82],"we":[83,113,145],"show":[84],"how":[85],"train":[87],"BERT-based":[89],"MWER":[93],"loss,":[94],"incorporate":[96],"improvements":[98],"of":[99,105],"loss":[102],"into":[103,122],"fine-tuning":[104],"deep":[106],"pretrained":[108,133],"models":[109],"propose":[114,137],"fusion":[116],"strategy":[117],"incorporates":[119],"MLM":[121],"process":[126],"effectively":[128],"distill":[129],"knowledge":[130],"model.":[134,197],"We":[135,166],"further":[136],"alternative":[139],"loss.":[141],"This":[142],"approach,":[143],"which":[144],"call":[146],"RescoreBERT,":[147],"reduces":[148,183],"6.6%/3.4%":[151],"relative":[152],"LibriSpeech":[155],"clean/other":[156],"test":[157],"sets":[158],"over":[159,193],"baseline":[162],"without":[163],"objective.":[165],"also":[167],"evaluate":[168],"our":[169],"method":[170],"internal":[173],"dataset":[174],"conversational":[177],"agent":[178],"and":[179,186],"find":[180],"both":[184],"latency":[185],"(by":[188],"3":[189],"8%":[191],"relative)":[192],"LSTM":[195]},"counts_by_year":[{"year":2025,"cited_by_count":7},{"year":2024,"cited_by_count":15},{"year":2023,"cited_by_count":10},{"year":2022,"cited_by_count":4}],"updated_date":"2026-03-10T16:38:18.471706","created_date":"2025-10-10T00:00:00"}