{"id":"https://openalex.org/W4319862674","doi":"https://doi.org/10.1109/slt54892.2023.10022622","title":"Residual Adapters for Targeted Updates in RNN-Transducer Based Speech Recognition System","display_name":"Residual Adapters for Targeted Updates in RNN-Transducer Based Speech Recognition System","publication_year":2023,"publication_date":"2023-01-09","ids":{"openalex":"https://openalex.org/W4319862674","doi":"https://doi.org/10.1109/slt54892.2023.10022622"},"language":"en","primary_location":{"id":"doi:10.1109/slt54892.2023.10022622","is_oa":false,"landing_page_url":"http://dx.doi.org/10.1109/slt54892.2023.10022622","pdf_url":null,"source":{"id":"https://openalex.org/S4363605953","display_name":"2022 IEEE Spoken Language Technology Workshop (SLT)","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":"2022 IEEE Spoken Language Technology Workshop (SLT)","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/A5052990279","display_name":"Sung\u2010Jun Han","orcid":null},"institutions":[{"id":"https://openalex.org/I100066346","display_name":"University of Stuttgart","ror":"https://ror.org/04vnq7t77","country_code":"DE","type":"education","lineage":["https://openalex.org/I100066346"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Sungjun Han","raw_affiliation_strings":["Universität Stuttgart,Germany"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Universität Stuttgart,Germany","institution_ids":["https://openalex.org/I100066346"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5067322015","display_name":"Deepak Baby","orcid":"https://orcid.org/0000-0001-8935-9068"},"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":"Deepak Baby","raw_affiliation_strings":["Amazon Alexa"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Amazon Alexa","institution_ids":["https://openalex.org/I1311688040"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5042701876","display_name":"Valentin Mendelev","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":"Valentin Mendelev","raw_affiliation_strings":["Amazon Alexa"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Amazon Alexa","institution_ids":["https://openalex.org/I1311688040"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.4356,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.54385937,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":94},"biblio":{"volume":"30","issue":null,"first_page":"160","last_page":"166"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9997000098228455,"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":0.9997000098228455,"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":0.9980000257492065,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.9980000257492065,"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":0.8130781650543213},{"id":"https://openalex.org/keywords/residual","display_name":"Residual","score":0.7398848533630371},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.7293336391448975},{"id":"https://openalex.org/keywords/adapter","display_name":"Adapter (computing)","score":0.6308097243309021},{"id":"https://openalex.org/keywords/recurrent-neural-network","display_name":"Recurrent neural network","score":0.5254096984863281},{"id":"https://openalex.org/keywords/dependency","display_name":"Dependency (UML)","score":0.5245749354362488},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.48228219151496887},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.44668444991111755},{"id":"https://openalex.org/keywords/recall","display_name":"Recall","score":0.44386523962020874},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.10479909181594849},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.09161791205406189}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8130781650543213},{"id":"https://openalex.org/C155512373","wikidata":"https://www.wikidata.org/wiki/Q287450","display_name":"Residual","level":2,"score":0.7398848533630371},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.7293336391448975},{"id":"https://openalex.org/C177284502","wikidata":"https://www.wikidata.org/wiki/Q1005390","display_name":"Adapter (computing)","level":2,"score":0.6308097243309021},{"id":"https://openalex.org/C147168706","wikidata":"https://www.wikidata.org/wiki/Q1457734","display_name":"Recurrent neural network","level":3,"score":0.5254096984863281},{"id":"https://openalex.org/C19768560","wikidata":"https://www.wikidata.org/wiki/Q320727","display_name":"Dependency (UML)","level":2,"score":0.5245749354362488},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.48228219151496887},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.44668444991111755},{"id":"https://openalex.org/C100660578","wikidata":"https://www.wikidata.org/wiki/Q18733","display_name":"Recall","level":2,"score":0.44386523962020874},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.10479909181594849},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.09161791205406189},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.0},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"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:10.1109/slt54892.2023.10022622","is_oa":false,"landing_page_url":"http://dx.doi.org/10.1109/slt54892.2023.10022622","pdf_url":null,"source":{"id":"https://openalex.org/S4363605953","display_name":"2022 IEEE Spoken Language Technology Workshop (SLT)","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":"2022 IEEE Spoken Language Technology Workshop (SLT)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4","score":0.49000000953674316}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":26,"referenced_works":["https://openalex.org/W1522301498","https://openalex.org/W2064675550","https://openalex.org/W2473930607","https://openalex.org/W2884282566","https://openalex.org/W2962760690","https://openalex.org/W2963211188","https://openalex.org/W2963240019","https://openalex.org/W2963250244","https://openalex.org/W2964303773","https://openalex.org/W2971840980","https://openalex.org/W3008898571","https://openalex.org/W3015194534","https://openalex.org/W3095311338","https://openalex.org/W3096815019","https://openalex.org/W3097777922","https://openalex.org/W3109311991","https://openalex.org/W3153675281","https://openalex.org/W3162244132","https://openalex.org/W3201225328","https://openalex.org/W3202576380","https://openalex.org/W3205949070","https://openalex.org/W4226316089","https://openalex.org/W4292595872","https://openalex.org/W4297841539","https://openalex.org/W6631190155","https://openalex.org/W6738045163"],"related_works":["https://openalex.org/W2133028525","https://openalex.org/W4306381730","https://openalex.org/W4229060448","https://openalex.org/W2981692913","https://openalex.org/W3184035966","https://openalex.org/W3044188621","https://openalex.org/W2160602540","https://openalex.org/W2485605994","https://openalex.org/W4385571108","https://openalex.org/W2146090693"],"abstract_inverted_index":{"This":[0],"paper":[1],"investigates":[2],"an":[3],"approach":[4],"for":[5,57],"adapting":[6],"RNN-Transducer":[7],"(RNN-T)":[8],"based":[9],"automatic":[10],"speech":[11],"recognition":[12,18],"(ASR)":[13],"model":[14,37,82],"to":[15,32,38,66,75,93,103,150],"improve":[16],"the":[17,35,51,58,80,95,101,113,118,122,125,135,146,151],"of":[19,42,70,124],"unseen":[20],"words":[21,106,137],"during":[22],"training.":[23],"Prior":[24],"works":[25],"have":[26],"shown":[27],"that":[28,98],"it":[29],"is":[30,54],"possible":[31],"incrementally":[33],"finetune":[34],"ASR":[36],"recognize":[39,104],"multiple":[40],"sets":[41],"new":[43,105,136],"words.":[44],"However,":[45],"this":[46],"creates":[47],"a":[48,109],"dependency":[49],"between":[50],"updates":[52],"which":[53,120],"not":[55],"ideal":[56],"hot-fixing":[59],"use-case":[60],"where":[61],"we":[62],"want":[63],"each":[64],"update":[65],"be":[67],"applied":[68],"independently":[69],"other":[71],"updates.":[72],"We":[73,89],"propose":[74],"train":[76],"residual":[77],"adapters":[78,96,126],"on":[79,112,134,145],"RNN-T":[81,153],"and":[83],"combine":[84,94],"them":[85],"on-the-fly":[86],"through":[87],"adapter-fusion.":[88],"investigate":[90],"several":[91],"approaches":[92],"so":[97],"they":[99],"maintain":[100],"ability":[102],"with":[107,138],"only":[108],"minimal":[110],"degradation":[111,144],"usual":[114,147],"user":[115],"requests.":[116],"Specifically,":[117],"sum-fusion":[119],"sums":[121],"outputs":[123],"inserted":[127],"in":[128],"parallel":[129],"shows":[130],"over":[131],"90%":[132],"recall":[133],"less":[139],"than":[140],"1%":[141],"relative":[142],"WER":[143],"data":[148],"compared":[149],"original":[152],"model.":[154]},"counts_by_year":[{"year":2024,"cited_by_count":1},{"year":2023,"cited_by_count":1}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}