{"id":"https://openalex.org/W4224916448","doi":"https://doi.org/10.1109/icassp43922.2022.9746801","title":"Knowledge Transfer from Large-Scale Pretrained Language Models to End-To-End Speech Recognizers","display_name":"Knowledge Transfer from Large-Scale Pretrained Language Models to End-To-End Speech Recognizers","publication_year":2022,"publication_date":"2022-04-27","ids":{"openalex":"https://openalex.org/W4224916448","doi":"https://doi.org/10.1109/icassp43922.2022.9746801"},"language":"en","primary_location":{"id":"doi:10.1109/icassp43922.2022.9746801","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp43922.2022.9746801","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":"conference-paper","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/A5102252420","display_name":"Yotaro Kubo","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":"Yotaro Kubo","raw_affiliation_strings":["Google,Tokyo","Google, Tokyo"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google,Tokyo","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google, Tokyo","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036532214","display_name":"Shigeki Karita","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":"Shigeki Karita","raw_affiliation_strings":["Google,Tokyo","Google, Tokyo"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google,Tokyo","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google, Tokyo","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5049614700","display_name":"Michiel Bacchiani","orcid":"https://orcid.org/0000-0003-4527-0197"},"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":"Michiel Bacchiani","raw_affiliation_strings":["Google,Tokyo","Google, Tokyo"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google,Tokyo","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google, Tokyo","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":null,"has_fulltext":false,"cited_by_count":18,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"8512","last_page":"8516"},"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9983999729156494,"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":0.9961000084877014,"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":0.8643866777420044},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.6951183676719666},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.6533284187316895},{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.649607241153717},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.6384234428405762},{"id":"https://openalex.org/keywords/end-to-end-principle","display_name":"End-to-end principle","score":0.5900524854660034},{"id":"https://openalex.org/keywords/constraint","display_name":"Constraint (computer-aided design)","score":0.5138075947761536},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4841678738594055},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.47532418370246887},{"id":"https://openalex.org/keywords/acoustic-model","display_name":"Acoustic model","score":0.4629247188568115},{"id":"https://openalex.org/keywords/reduction","display_name":"Reduction (mathematics)","score":0.45150575041770935},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.4315491020679474},{"id":"https://openalex.org/keywords/speech-processing","display_name":"Speech processing","score":0.2944720983505249},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.08880513906478882}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8643866777420044},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.6951183676719666},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.6533284187316895},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.649607241153717},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.6384234428405762},{"id":"https://openalex.org/C74296488","wikidata":"https://www.wikidata.org/wiki/Q2527392","display_name":"End-to-end principle","level":2,"score":0.5900524854660034},{"id":"https://openalex.org/C2776036281","wikidata":"https://www.wikidata.org/wiki/Q48769818","display_name":"Constraint (computer-aided design)","level":2,"score":0.5138075947761536},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4841678738594055},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.47532418370246887},{"id":"https://openalex.org/C155635449","wikidata":"https://www.wikidata.org/wiki/Q4674699","display_name":"Acoustic model","level":3,"score":0.4629247188568115},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.45150575041770935},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4315491020679474},{"id":"https://openalex.org/C61328038","wikidata":"https://www.wikidata.org/wiki/Q3358061","display_name":"Speech processing","level":2,"score":0.2944720983505249},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.08880513906478882},{"id":"https://openalex.org/C78519656","wikidata":"https://www.wikidata.org/wiki/Q101333","display_name":"Mechanical engineering","level":1,"score":0.0},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.0},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","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}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp43922.2022.9746801","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp43922.2022.9746801","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"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education","score":0.5099999904632568}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":27,"referenced_works":["https://openalex.org/W1494198834","https://openalex.org/W1522301498","https://openalex.org/W1566289585","https://openalex.org/W1821462560","https://openalex.org/W1828163288","https://openalex.org/W2121879602","https://openalex.org/W2327501763","https://openalex.org/W2842511635","https://openalex.org/W2896457183","https://openalex.org/W2936774411","https://openalex.org/W2962760690","https://openalex.org/W2963362078","https://openalex.org/W2963382687","https://openalex.org/W3036601975","https://openalex.org/W3093579165","https://openalex.org/W3096297644","https://openalex.org/W3096518646","https://openalex.org/W3097777922","https://openalex.org/W3162649911","https://openalex.org/W3164692279","https://openalex.org/W4297808394","https://openalex.org/W6631190155","https://openalex.org/W6638523607","https://openalex.org/W6638749077","https://openalex.org/W6755207826","https://openalex.org/W6780218876","https://openalex.org/W6784614252"],"related_works":["https://openalex.org/W2359001871","https://openalex.org/W3015923889","https://openalex.org/W2916997151","https://openalex.org/W62743518","https://openalex.org/W2949174760","https://openalex.org/W2998429374","https://openalex.org/W2782005958","https://openalex.org/W2963414781","https://openalex.org/W3112480982","https://openalex.org/W4360942793"],"abstract_inverted_index":{"End-to-end":[0],"speech":[1,11,36],"recognition":[2,12],"is":[3,55,62],"a":[4,25,31,74,84],"promising":[5],"technology":[6],"for":[7,76,140,174],"enabling":[8],"compact":[9],"automatic":[10],"(ASR)":[13],"systems":[14,168],"since":[15],"it":[16],"can":[17,65,90,115],"unify":[18],"the":[19,184],"acoustic":[20],"and":[21,64,128,153],"language":[22,85,110],"model":[23,86],"into":[24],"single":[26],"neural":[27,87,154],"network.":[28],"However,":[29],"as":[30,118,125],"drawback,":[32],"training":[33,158],"of":[34,108,121,148],"end-to-end":[35,43],"recognizers":[37],"always":[38],"requires":[39],"transcribed":[40,60],"utterances.":[41],"Since":[42,112],"models":[44],"are":[45,132],"also":[46,133],"known":[47],"to":[48,100,135,161,171],"be":[49,67,91,116,136,172],"severely":[50],"data":[51],"hungry,":[52],"this":[53,78,97],"constraint":[54],"crucial":[56],"especially":[57],"because":[58],"obtaining":[59],"utterances":[61],"costly":[63],"possibly":[66],"impractical":[68],"or":[69],"impossible.":[70],"This":[71,143],"paper":[72,98,144],"proposes":[73],"method":[75],"alleviating":[77],"issue":[79],"by":[80,156],"transferring":[81],"knowledge":[82,103],"from":[83],"network":[88],"that":[89],"pretrained":[92],"with":[93],"text-only":[94],"data.":[95],"Specifically,":[96],"attempts":[99],"transfer":[101],"semantic":[102],"acquired":[104],"in":[105,183],"embedding":[106,113,163],"vectors":[107,114],"large-scale":[109],"models.":[111],"assumed":[117],"implicit":[119],"representations":[120],"linguistic":[122],"information":[123],"such":[124],"part-of-speech,":[126],"intent,":[127],"so":[129],"on,":[130],"those":[131],"expected":[134],"useful":[137],"modeling":[138],"cues":[139],"ASR":[141,149],"decoders.":[142],"extends":[145],"two":[146],"types":[147],"decoders,":[150],"attention-based":[151],"decoders":[152],"transducers,":[155],"modifying":[157],"loss":[159],"functions":[160],"include":[162],"prediction":[164],"terms.":[165],"The":[166],"proposed":[167],"were":[169],"shown":[170],"effective":[173],"error":[175],"rate":[176],"reduction":[177],"without":[178],"incurring":[179],"extra":[180],"computational":[181],"costs":[182],"decoding":[185],"phase.":[186]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2024,"cited_by_count":8},{"year":2023,"cited_by_count":8},{"year":2022,"cited_by_count":1}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}