{"id":"https://openalex.org/W3205788551","doi":"https://doi.org/10.21437/interspeech.2022-13","title":"Internal Language Model Adaptation with Text-Only Data for End-to-End Speech Recognition","display_name":"Internal Language Model Adaptation with Text-Only Data for End-to-End Speech Recognition","publication_year":2022,"publication_date":"2022-09-16","ids":{"openalex":"https://openalex.org/W3205788551","doi":"https://doi.org/10.21437/interspeech.2022-13","mag":"3205788551"},"language":"en","primary_location":{"id":"doi:10.21437/interspeech.2022-13","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2022-13","pdf_url":null,"source":{"id":"https://openalex.org/S4363604309","display_name":"Interspeech 2022","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":"Interspeech 2022","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/A5101749753","display_name":"Zhong Meng","orcid":"https://orcid.org/0000-0001-7814-5929"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhong Meng","raw_affiliation_strings":["Microsoft Corporation, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Corporation, USA","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034136587","display_name":"Yashesh Gaur","orcid":null},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yashesh Gaur","raw_affiliation_strings":["Microsoft Corporation, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Corporation, USA","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5016279564","display_name":"Naoyuki Kanda","orcid":"https://orcid.org/0000-0002-8628-3288"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Naoyuki Kanda","raw_affiliation_strings":["Microsoft Corporation, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Corporation, USA","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100365053","display_name":"Jinyu Li","orcid":"https://orcid.org/0000-0002-1089-9748"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jinyu Li","raw_affiliation_strings":["Microsoft Corporation, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Corporation, USA","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100329117","display_name":"Xie Chen","orcid":"https://orcid.org/0000-0001-5801-2571"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xie Chen","raw_affiliation_strings":["Microsoft Corporation, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Corporation, USA","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101709477","display_name":"Yu Wu","orcid":"https://orcid.org/0000-0002-5715-3011"},"institutions":[{"id":"https://openalex.org/I4210113369","display_name":"Microsoft Research Asia (China)","ror":"https://ror.org/0300m5276","country_code":"CN","type":"company","lineage":["https://openalex.org/I1290206253","https://openalex.org/I4210113369"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yu Wu","raw_affiliation_strings":["Microsoft Research Asia, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Research Asia, China","institution_ids":["https://openalex.org/I4210113369"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5077401426","display_name":"Yifan Gong","orcid":"https://orcid.org/0000-0001-8786-3391"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yifan Gong","raw_affiliation_strings":["Microsoft Corporation, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Corporation, USA","institution_ids":["https://openalex.org/I1290206253"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":7,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.9725,"has_fulltext":false,"cited_by_count":19,"citation_normalized_percentile":{"value":0.88067851,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":99},"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"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":0.9980999827384949,"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/T10860","display_name":"Speech and Audio Processing","score":0.9970999956130981,"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":0.7886376976966858},{"id":"https://openalex.org/keywords/end-to-end-principle","display_name":"End-to-end principle","score":0.7544207572937012},{"id":"https://openalex.org/keywords/adaptation","display_name":"Adaptation (eye)","score":0.6505216360092163},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5905680060386658},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.48555490374565125},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.4539688229560852},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.41666683554649353},{"id":"https://openalex.org/keywords/linguistics","display_name":"Linguistics","score":0.37774014472961426},{"id":"https://openalex.org/keywords/psychology","display_name":"Psychology","score":0.16549983620643616}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7886376976966858},{"id":"https://openalex.org/C74296488","wikidata":"https://www.wikidata.org/wiki/Q2527392","display_name":"End-to-end principle","level":2,"score":0.7544207572937012},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.6505216360092163},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5905680060386658},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.48555490374565125},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4539688229560852},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.41666683554649353},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.37774014472961426},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.16549983620643616},{"id":"https://openalex.org/C169760540","wikidata":"https://www.wikidata.org/wiki/Q207011","display_name":"Neuroscience","level":1,"score":0.0},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.21437/interspeech.2022-13","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2022-13","pdf_url":null,"source":{"id":"https://openalex.org/S4363604309","display_name":"Interspeech 2022","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":"Interspeech 2022","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":64,"referenced_works":["https://openalex.org/W179875071","https://openalex.org/W854541894","https://openalex.org/W1513820424","https://openalex.org/W1828163288","https://openalex.org/W1915251500","https://openalex.org/W1922655562","https://openalex.org/W1965555277","https://openalex.org/W1989549063","https://openalex.org/W2010362084","https://openalex.org/W2080005694","https://openalex.org/W2094147890","https://openalex.org/W2095705004","https://openalex.org/W2127141656","https://openalex.org/W2293634267","https://openalex.org/W2402040300","https://openalex.org/W2510867321","https://openalex.org/W2515801922","https://openalex.org/W2584667682","https://openalex.org/W2636483419","https://openalex.org/W2653985929","https://openalex.org/W2769025471","https://openalex.org/W2795867901","https://openalex.org/W2796339975","https://openalex.org/W2888779557","https://openalex.org/W2891980359","https://openalex.org/W2892095314","https://openalex.org/W2911629330","https://openalex.org/W2936252403","https://openalex.org/W2939164678","https://openalex.org/W2943008652","https://openalex.org/W2962784628","https://openalex.org/W2962824709","https://openalex.org/W2963174344","https://openalex.org/W2963362078","https://openalex.org/W2963403868","https://openalex.org/W2963970535","https://openalex.org/W2972621414","https://openalex.org/W2972818416","https://openalex.org/W2990501692","https://openalex.org/W3008008574","https://openalex.org/W3008037978","https://openalex.org/W3011956499","https://openalex.org/W3015194534","https://openalex.org/W3015915933","https://openalex.org/W3016010032","https://openalex.org/W3016234571","https://openalex.org/W3080248383","https://openalex.org/W3093620947","https://openalex.org/W3094667432","https://openalex.org/W3095311338","https://openalex.org/W3095697114","https://openalex.org/W3096815019","https://openalex.org/W3152221657","https://openalex.org/W3153201663","https://openalex.org/W3161873870","https://openalex.org/W3163560333","https://openalex.org/W3184976814","https://openalex.org/W3197507772","https://openalex.org/W3197661863","https://openalex.org/W3198439131","https://openalex.org/W4286784498","https://openalex.org/W4288072840","https://openalex.org/W4303494751","https://openalex.org/W4385245566"],"related_works":["https://openalex.org/W3179968364","https://openalex.org/W3196421258","https://openalex.org/W2151749779","https://openalex.org/W2997567050","https://openalex.org/W2938107654","https://openalex.org/W3008587939","https://openalex.org/W3127005940","https://openalex.org/W1999612375","https://openalex.org/W2916997151","https://openalex.org/W2949174760"],"abstract_inverted_index":{"Text-only":[0],"adaptation":[1,39,159],"of":[2,41,132,151,160],"an":[3,21,36,51,56,109],"end-to-end":[4],"(E2E)":[5],"model":[6,16,44,53,71,107,163],"remains":[7],"a":[8,94,156],"challenging":[9],"task":[10],"for":[11],"automatic":[12],"speech":[13],"recognition":[14],"(ASR).Language":[15],"(LM)":[17],"fusion-based":[18],"approaches":[19],"require":[20],"additional":[22],"external":[23],"LM":[24,38,58,111],"during":[25],"inference,":[26],"significantly":[27],"increasing":[28,165],"the":[29,42,61,69,76,82,86,90,105,114,125,129,133,140,147,152,161,166,187],"computation":[30],"cost.To":[31],"overcome":[32],"this,":[33],"we":[34,80,118,144],"propose":[35,119],"internal":[37,57,83,110,137],"(ILMA)":[40],"E2E":[43,52,70,87,106,116,162],"using":[45],"text-only":[46,158],"data.Trained":[47],"with":[48,108,170],"audio-transcript":[49],"pairs,":[50],"implicitly":[54],"learns":[55],"that":[59],"characterizes":[60],"token":[62],"sequence":[63],"probability":[64],"which":[65],"is":[66,101,139],"approximated":[67],"by":[68,123],"output":[72,130],"after":[73],"zeroing":[74],"out":[75],"encoder":[77],"contribution.During":[78],"ILMA,":[79],"fine-tune":[81],"LM,":[84],"i.e.,":[85],"components":[88],"excluding":[89],"encoder,":[91],"to":[92,103,120,179],"minimize":[93],"cross-entropy":[95],"loss.To":[96],"make":[97],"ILMA":[98,122,176],"effective,":[99],"it":[100],"essential":[102],"train":[104],"loss":[112],"besides":[113],"standard":[115],"loss.Furthermore,":[117],"regularize":[121],"minimizing":[124],"Kullback-Leibler":[126],"divergence":[127],"between":[128],"distributions":[131],"adapted":[134],"and":[135],"unadapted":[136,188],"LMs.ILMA":[138],"most":[141],"effective":[142],"when":[143],"update":[145],"only":[146],"last":[148],"linear":[149],"layer":[150],"joint":[153],"network.ILMA":[154],"enables":[155],"fast":[157],"without":[164],"run-time":[167],"computational":[168],"cost.Experimented":[169],"30K-hour":[171],"trained":[172],"transformer":[173],"transducer":[174],"models,":[175],"achieves":[177],"up":[178],"34.9%":[180],"relative":[181],"word":[182],"error":[183],"rate":[184],"reduction":[185],"from":[186],"baseline.":[189]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2024,"cited_by_count":5},{"year":2023,"cited_by_count":11},{"year":2022,"cited_by_count":2}],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}