{"id":"https://openalex.org/W4375868951","doi":"https://doi.org/10.1109/icassp49357.2023.10094997","title":"Fine-Grained Textual Knowledge Transfer to Improve RNN Transducers for Speech Recognition and Understanding","display_name":"Fine-Grained Textual Knowledge Transfer to Improve RNN Transducers for Speech Recognition and Understanding","publication_year":2023,"publication_date":"2023-05-05","ids":{"openalex":"https://openalex.org/W4375868951","doi":"https://doi.org/10.1109/icassp49357.2023.10094997"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49357.2023.10094997","is_oa":false,"landing_page_url":"http://dx.doi.org/10.1109/icassp49357.2023.10094997","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2023 - 2023 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/A5113145191","display_name":"Vishal Sunder","orcid":null},"institutions":[{"id":"https://openalex.org/I52357470","display_name":"The Ohio State University","ror":"https://ror.org/00rs6vg23","country_code":"US","type":"education","lineage":["https://openalex.org/I52357470"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Vishal Sunder","raw_affiliation_strings":["The Ohio State University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Ohio State University","institution_ids":["https://openalex.org/I52357470"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101787514","display_name":"Samuel Thomas","orcid":"https://orcid.org/0000-0001-7573-0620"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Samuel Thomas","raw_affiliation_strings":["IBM Research AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"IBM Research AI","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5110252428","display_name":"Hong-Kwang Jeff Kuo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hong-Kwang J. Kuo","raw_affiliation_strings":["IBM Research AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"IBM Research AI","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5003725957","display_name":"Brian Kingsbury","orcid":"https://orcid.org/0000-0002-1343-6837"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Brian Kingsbury","raw_affiliation_strings":["IBM Research AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"IBM Research AI","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5056667180","display_name":"Eric Fosler\u2010Lussier","orcid":"https://orcid.org/0000-0001-8004-5169"},"institutions":[{"id":"https://openalex.org/I52357470","display_name":"The Ohio State University","ror":"https://ror.org/00rs6vg23","country_code":"US","type":"education","lineage":["https://openalex.org/I52357470"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Eric Fosler-Lussier","raw_affiliation_strings":["The Ohio State University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Ohio State University","institution_ids":["https://openalex.org/I52357470"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":6,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"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/T10860","display_name":"Speech and Audio Processing","score":0.9976999759674072,"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/T12031","display_name":"Speech and dialogue systems","score":0.9966999888420105,"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":0.8861386775970459},{"id":"https://openalex.org/keywords/nist","display_name":"NIST","score":0.7626975774765015},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.7572444677352905},{"id":"https://openalex.org/keywords/recurrent-neural-network","display_name":"Recurrent neural network","score":0.6722270250320435},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.6483526229858398},{"id":"https://openalex.org/keywords/transfer-of-learning","display_name":"Transfer of learning","score":0.630695104598999},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5890553593635559},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5083605647087097},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.5068071484565735},{"id":"https://openalex.org/keywords/acoustic-model","display_name":"Acoustic model","score":0.461723268032074},{"id":"https://openalex.org/keywords/spoken-language","display_name":"Spoken language","score":0.45760655403137207},{"id":"https://openalex.org/keywords/knowledge-base","display_name":"Knowledge base","score":0.4249325394630432},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.4134131073951721},{"id":"https://openalex.org/keywords/speech-processing","display_name":"Speech processing","score":0.2696085572242737},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.12019714713096619}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8861386775970459},{"id":"https://openalex.org/C111219384","wikidata":"https://www.wikidata.org/wiki/Q6954384","display_name":"NIST","level":2,"score":0.7626975774765015},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.7572444677352905},{"id":"https://openalex.org/C147168706","wikidata":"https://www.wikidata.org/wiki/Q1457734","display_name":"Recurrent neural network","level":3,"score":0.6722270250320435},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.6483526229858398},{"id":"https://openalex.org/C150899416","wikidata":"https://www.wikidata.org/wiki/Q1820378","display_name":"Transfer of learning","level":2,"score":0.630695104598999},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5890553593635559},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5083605647087097},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5068071484565735},{"id":"https://openalex.org/C155635449","wikidata":"https://www.wikidata.org/wiki/Q4674699","display_name":"Acoustic model","level":3,"score":0.461723268032074},{"id":"https://openalex.org/C2776230583","wikidata":"https://www.wikidata.org/wiki/Q1322198","display_name":"Spoken language","level":2,"score":0.45760655403137207},{"id":"https://openalex.org/C4554734","wikidata":"https://www.wikidata.org/wiki/Q593744","display_name":"Knowledge base","level":2,"score":0.4249325394630432},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4134131073951721},{"id":"https://openalex.org/C61328038","wikidata":"https://www.wikidata.org/wiki/Q3358061","display_name":"Speech processing","level":2,"score":0.2696085572242737},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.12019714713096619},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C187736073","wikidata":"https://www.wikidata.org/wiki/Q2920921","display_name":"Management","level":1,"score":0.0},{"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/icassp49357.2023.10094997","is_oa":false,"landing_page_url":"http://dx.doi.org/10.1109/icassp49357.2023.10094997","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","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":28,"referenced_works":["https://openalex.org/W1828163288","https://openalex.org/W2407080277","https://openalex.org/W2896457183","https://openalex.org/W2908510526","https://openalex.org/W2936774411","https://openalex.org/W2937780860","https://openalex.org/W3016006013","https://openalex.org/W3096297644","https://openalex.org/W3096589040","https://openalex.org/W3097964672","https://openalex.org/W3098169801","https://openalex.org/W3100460087","https://openalex.org/W3148001440","https://openalex.org/W3155162503","https://openalex.org/W3161000786","https://openalex.org/W3163300396","https://openalex.org/W3167533889","https://openalex.org/W3209371554","https://openalex.org/W3211224152","https://openalex.org/W4224916448","https://openalex.org/W4226315417","https://openalex.org/W4296068593","https://openalex.org/W4296068780","https://openalex.org/W4296070421","https://openalex.org/W4300900294","https://openalex.org/W4385567350","https://openalex.org/W6755207826","https://openalex.org/W6792221667"],"related_works":["https://openalex.org/W2160451571","https://openalex.org/W2495256954","https://openalex.org/W2126322296","https://openalex.org/W2890297197","https://openalex.org/W2955724459","https://openalex.org/W2037769829","https://openalex.org/W2163537793","https://openalex.org/W2916997151","https://openalex.org/W2781555308","https://openalex.org/W2402899696"],"abstract_inverted_index":{"RNN":[0],"Tranducer":[1],"(RNN-T)":[2],"technology":[3],"is":[4,113,188],"very":[5],"popular":[6],"for":[7,11,62,106,120],"building":[8],"deployable":[9],"models":[10,27,44,183],"end-to-end":[12],"(E2E)":[13],"automatic":[14],"speech":[15,30,182,198],"recognition":[16],"(ASR)":[17],"and":[18,67,131,141,170,192],"spoken":[19],"language":[20,50],"understanding":[21,51],"(SLU).":[22],"Since":[23],"these":[24,74],"are":[25],"E2E":[26,65],"operating":[28],"on":[29,128,142,148,162],"directly,":[31],"there":[32],"remains":[33],"a":[34,58,90,152,172],"potential":[35],"to":[36,69,175,178],"improve":[37,85,125],"their":[38],"performance":[39,87,127,173],"using":[40,89],"purely":[41],"text":[42],"based":[43,64,181],"like":[45,184],"BERT,":[46],"which":[47,149],"have":[48],"strong":[49],"capabilities.":[52],"In":[53,76,97],"this":[54],"paper,":[55],"we":[56,84,101,150,158],"propose":[57],"new":[59,153],"training":[60],"criteria":[61],"RNN-T":[63,118],"ASR":[66,86,104,126],"SLU":[68,107],"transfer":[70,94],"BERT\u2019s":[71],"knowledge":[72,93,112],"into":[73],"systems.":[75],"the":[77,98,103,110,117,129,136,143,163],"first":[78],"stage":[79],"of":[80,135,197],"our":[81,186],"proposed":[82],"mechanism,":[83],"by":[88,116],"fine-grained,":[91],"tokenwise":[92],"from":[95],"BERT.":[96],"second":[99],"stage,":[100],"fine-tune":[102],"model":[105,119,187],"such":[108],"that":[109],"above":[111],"explicitly":[114],"utilized":[115],"improved":[121],"performance.":[122,155],"Our":[123],"techniques":[124],"Switchboard":[130],"CallHome":[132],"test":[133],"sets":[134],"NIST":[137],"Hub5":[138],"2000":[139],"evaluation":[140],"recently":[144],"released":[145],"SLURP":[146,164],"dataset":[147],"achieve":[151],"state-of-the-art":[154],"For":[156],"SLU,":[157],"show":[159],"significant":[160],"improvements":[161],"slot":[165],"filling":[166],"task,":[167],"outperforming":[168],"HuBERT-base":[169],"reaching":[171],"close":[174],"HuBERTlarge.":[176],"Compared":[177],"large":[179],"transformer":[180],"HuBERT,":[185],"significantly":[189],"more":[190],"compact":[191],"uses":[193],"only":[194],"300":[195],"hours":[196],"pretraining":[199],"data.":[200]},"counts_by_year":[{"year":2024,"cited_by_count":4},{"year":2023,"cited_by_count":2}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}