{"id":"https://openalex.org/W4221151615","doi":"https://doi.org/10.1109/icassp43922.2022.9747418","title":"Improving the Fusion of Acoustic and Text Representations in RNN-T","display_name":"Improving the Fusion of Acoustic and Text Representations in RNN-T","publication_year":2022,"publication_date":"2022-04-27","ids":{"openalex":"https://openalex.org/W4221151615","doi":"https://doi.org/10.1109/icassp43922.2022.9747418"},"language":"en","primary_location":{"id":"doi:10.1109/icassp43922.2022.9747418","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp43922.2022.9747418","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/A5100460139","display_name":"Chao Zhang","orcid":"https://orcid.org/0000-0002-1018-4144"},"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":"Chao Zhang","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100677404","display_name":"Bo Li","orcid":"https://orcid.org/0000-0003-1483-8077"},"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":"Bo Li","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089287882","display_name":"Zhiyun Lu","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":"Zhiyun Lu","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","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 LLC,USA","Google LLC, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5001306222","display_name":"Shuo-Yiin Chang","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":"Shuo-yiin Chang","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","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":9,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"8117","last_page":"8121"},"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/T11309","display_name":"Music and Audio Processing","score":0.9994000196456909,"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":0.9988999962806702,"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/recurrent-neural-network","display_name":"Recurrent neural network","score":0.8529341816902161},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7920759320259094},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.7107110023498535},{"id":"https://openalex.org/keywords/pooling","display_name":"Pooling","score":0.7046124935150146},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.6579182147979736},{"id":"https://openalex.org/keywords/word-error-rate","display_name":"Word error rate","score":0.4774094820022583},{"id":"https://openalex.org/keywords/layer","display_name":"Layer (electronics)","score":0.4499203562736511},{"id":"https://openalex.org/keywords/word","display_name":"Word (group theory)","score":0.43761318922042847},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4354187548160553},{"id":"https://openalex.org/keywords/bilinear-interpolation","display_name":"Bilinear interpolation","score":0.4250972867012024},{"id":"https://openalex.org/keywords/joint","display_name":"Joint (building)","score":0.4109296500682831},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.3781405985355377},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.08933070302009583},{"id":"https://openalex.org/keywords/computer-vision","display_name":"Computer vision","score":0.07565602660179138}],"concepts":[{"id":"https://openalex.org/C147168706","wikidata":"https://www.wikidata.org/wiki/Q1457734","display_name":"Recurrent neural network","level":3,"score":0.8529341816902161},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7920759320259094},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.7107110023498535},{"id":"https://openalex.org/C70437156","wikidata":"https://www.wikidata.org/wiki/Q7228652","display_name":"Pooling","level":2,"score":0.7046124935150146},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.6579182147979736},{"id":"https://openalex.org/C40969351","wikidata":"https://www.wikidata.org/wiki/Q3516228","display_name":"Word error rate","level":2,"score":0.4774094820022583},{"id":"https://openalex.org/C2779227376","wikidata":"https://www.wikidata.org/wiki/Q6505497","display_name":"Layer (electronics)","level":2,"score":0.4499203562736511},{"id":"https://openalex.org/C90805587","wikidata":"https://www.wikidata.org/wiki/Q10944557","display_name":"Word (group theory)","level":2,"score":0.43761318922042847},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4354187548160553},{"id":"https://openalex.org/C205203396","wikidata":"https://www.wikidata.org/wiki/Q612143","display_name":"Bilinear interpolation","level":2,"score":0.4250972867012024},{"id":"https://openalex.org/C18555067","wikidata":"https://www.wikidata.org/wiki/Q8375051","display_name":"Joint (building)","level":2,"score":0.4109296500682831},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.3781405985355377},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.08933070302009583},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.07565602660179138},{"id":"https://openalex.org/C178790620","wikidata":"https://www.wikidata.org/wiki/Q11351","display_name":"Organic chemistry","level":1,"score":0.0},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.0},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","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/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0},{"id":"https://openalex.org/C170154142","wikidata":"https://www.wikidata.org/wiki/Q150737","display_name":"Architectural engineering","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp43922.2022.9747418","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp43922.2022.9747418","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":[{"score":0.699999988079071,"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":53,"referenced_works":["https://openalex.org/W854541894","https://openalex.org/W1508165687","https://openalex.org/W1828163288","https://openalex.org/W1942035323","https://openalex.org/W2064675550","https://openalex.org/W2105880865","https://openalex.org/W2121879602","https://openalex.org/W2127141656","https://openalex.org/W2143612262","https://openalex.org/W2170653751","https://openalex.org/W2327501763","https://openalex.org/W2526425061","https://openalex.org/W2750499125","https://openalex.org/W2928941594","https://openalex.org/W2935756939","https://openalex.org/W2949197413","https://openalex.org/W2962760690","https://openalex.org/W2963414781","https://openalex.org/W2963827914","https://openalex.org/W2964084166","https://openalex.org/W2985076077","https://openalex.org/W3007227084","https://openalex.org/W3008037978","https://openalex.org/W3015190365","https://openalex.org/W3015686596","https://openalex.org/W3015927303","https://openalex.org/W3015995734","https://openalex.org/W3016010032","https://openalex.org/W3016234571","https://openalex.org/W3028545098","https://openalex.org/W3093501662","https://openalex.org/W3094667432","https://openalex.org/W3097777922","https://openalex.org/W3160405885","https://openalex.org/W3160628828","https://openalex.org/W3161873870","https://openalex.org/W3162244132","https://openalex.org/W3162665866","https://openalex.org/W3163203022","https://openalex.org/W3163300396","https://openalex.org/W3163560333","https://openalex.org/W3198004110","https://openalex.org/W4210463634","https://openalex.org/W4226462878","https://openalex.org/W4289421997","https://openalex.org/W6623517193","https://openalex.org/W6638749077","https://openalex.org/W6675992320","https://openalex.org/W6728881024","https://openalex.org/W6747398299","https://openalex.org/W6754733129","https://openalex.org/W6760633627","https://openalex.org/W6784418856"],"related_works":["https://openalex.org/W2950524887","https://openalex.org/W2883502031","https://openalex.org/W2261271299","https://openalex.org/W4280638452","https://openalex.org/W4285020665","https://openalex.org/W2963066927","https://openalex.org/W3160506688","https://openalex.org/W3111811104","https://openalex.org/W4380083739","https://openalex.org/W2964944724"],"abstract_inverted_index":{"The":[0],"recurrent":[1],"neural":[2],"network":[3,36,55,80,112],"transducer":[4],"(RNN-T)":[5],"has":[6],"recently":[7],"become":[8],"the":[9,21,34,39,44,48,53,58,78,89,106,110,114,134,138],"mainstream":[10],"end-to-end":[11],"approach":[12],"for":[13,126],"streaming":[14],"automatic":[15],"speech":[16],"recognition":[17],"(ASR).":[18],"To":[19],"estimate":[20],"output":[22,90],"distributions":[23],"over":[24,129],"subword":[25,60],"units,":[26],"RNN-T":[27,117],"uses":[28],"a":[29,73,122,152],"fully":[30],"connected":[31],"layer":[32],"as":[33],"joint":[35,79,135],"to":[37,67,81,86,98],"fuse":[38],"acoustic":[40,45,101],"representations":[41,50,85],"extracted":[42],"using":[43,52],"encoder":[46,102],"with":[47,150],"text":[49],"obtained":[51],"prediction":[54,111],"based":[56],"on":[57,121],"previous":[59],"units.":[61],"In":[62],"this":[63],"paper,":[64],"we":[65],"propose":[66],"use":[68,136],"gating,":[69],"bilinear":[70],"pooling,":[71],"and":[72],"combination":[74],"of":[75,116,137],"them":[76],"in":[77,143],"produce":[82],"more":[83],"expressive":[84],"feed":[87],"into":[88,109],"layer.":[91],"A":[92],"regularisation":[93],"method":[94],"is":[95],"also":[96],"proposed":[97,139],"enable":[99],"better":[100],"training":[103],"by":[104],"reducing":[105],"gradients":[107],"back-propagated":[108],"at":[113],"beginning":[115],"training.":[118],"Experimental":[119],"results":[120],"multilingual":[123],"ASR":[124],"setting":[125],"voice":[127],"search":[128],"nine":[130],"languages":[131],"show":[132],"that":[133],"methods":[140],"can":[141],"result":[142],"4%\u20135%":[144],"relative":[145],"word":[146],"error":[147],"rate":[148],"reductions":[149],"only":[151],"few":[153],"million":[154],"extra":[155],"parameters.":[156]},"counts_by_year":[{"year":2024,"cited_by_count":2},{"year":2023,"cited_by_count":4},{"year":2022,"cited_by_count":3}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}