{"id":"https://openalex.org/W3140235797","doi":"https://doi.org/10.1109/slt48900.2021.9383560","title":"Deep Shallow Fusion for RNN-T Personalization","display_name":"Deep Shallow Fusion for RNN-T Personalization","publication_year":2021,"publication_date":"2021-01-19","ids":{"openalex":"https://openalex.org/W3140235797","doi":"https://doi.org/10.1109/slt48900.2021.9383560","mag":"3140235797"},"language":"en","primary_location":{"id":"doi:10.1109/slt48900.2021.9383560","is_oa":false,"landing_page_url":"https://doi.org/10.1109/slt48900.2021.9383560","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2021 IEEE Spoken Language Technology Workshop (SLT)","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/A5103098852","display_name":"Duc Le","orcid":"https://orcid.org/0000-0001-9490-2563"},"institutions":[{"id":"https://openalex.org/I2252078561","display_name":"Meta (Israel)","ror":"https://ror.org/02388em19","country_code":"IL","type":"company","lineage":["https://openalex.org/I2252078561","https://openalex.org/I4210114444"]}],"countries":["IL"],"is_corresponding":true,"raw_author_name":"Duc Le","raw_affiliation_strings":["Facebook AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Facebook AI","institution_ids":["https://openalex.org/I2252078561"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5048538280","display_name":"Gil Keren","orcid":"https://orcid.org/0000-0002-5153-3494"},"institutions":[{"id":"https://openalex.org/I2252078561","display_name":"Meta (Israel)","ror":"https://ror.org/02388em19","country_code":"IL","type":"company","lineage":["https://openalex.org/I2252078561","https://openalex.org/I4210114444"]}],"countries":["IL"],"is_corresponding":false,"raw_author_name":"Gil Keren","raw_affiliation_strings":["Facebook AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Facebook AI","institution_ids":["https://openalex.org/I2252078561"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5051075420","display_name":"Julian Chan","orcid":null},"institutions":[{"id":"https://openalex.org/I2252078561","display_name":"Meta (Israel)","ror":"https://ror.org/02388em19","country_code":"IL","type":"company","lineage":["https://openalex.org/I2252078561","https://openalex.org/I4210114444"]}],"countries":["IL"],"is_corresponding":false,"raw_author_name":"Julian Chan","raw_affiliation_strings":["Facebook AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Facebook AI","institution_ids":["https://openalex.org/I2252078561"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5074237839","display_name":"Jay Mahadeokar","orcid":null},"institutions":[{"id":"https://openalex.org/I2252078561","display_name":"Meta (Israel)","ror":"https://ror.org/02388em19","country_code":"IL","type":"company","lineage":["https://openalex.org/I2252078561","https://openalex.org/I4210114444"]}],"countries":["IL"],"is_corresponding":false,"raw_author_name":"Jay Mahadeokar","raw_affiliation_strings":["Facebook AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Facebook AI","institution_ids":["https://openalex.org/I2252078561"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5047073253","display_name":"Christian Fuegen","orcid":null},"institutions":[{"id":"https://openalex.org/I2252078561","display_name":"Meta (Israel)","ror":"https://ror.org/02388em19","country_code":"IL","type":"company","lineage":["https://openalex.org/I2252078561","https://openalex.org/I4210114444"]}],"countries":["IL"],"is_corresponding":false,"raw_author_name":"Christian Fuegen","raw_affiliation_strings":["Facebook AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Facebook AI","institution_ids":["https://openalex.org/I2252078561"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5041313589","display_name":"Michael L. Seltzer","orcid":"https://orcid.org/0000-0003-3474-2451"},"institutions":[{"id":"https://openalex.org/I2252078561","display_name":"Meta (Israel)","ror":"https://ror.org/02388em19","country_code":"IL","type":"company","lineage":["https://openalex.org/I2252078561","https://openalex.org/I4210114444"]}],"countries":["IL"],"is_corresponding":false,"raw_author_name":"Michael L. Seltzer","raw_affiliation_strings":["Facebook AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Facebook AI","institution_ids":["https://openalex.org/I2252078561"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5103098852"],"corresponding_institution_ids":["https://openalex.org/I2252078561"],"apc_list":null,"apc_paid":null,"fwci":7.2756,"has_fulltext":false,"cited_by_count":65,"citation_normalized_percentile":{"value":0.97562373,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":98,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"251","last_page":"257"},"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9998000264167786,"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":0.9991000294685364,"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/recurrent-neural-network","display_name":"Recurrent neural network","score":0.8061233162879944},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8016386032104492},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.6416541934013367},{"id":"https://openalex.org/keywords/personalization","display_name":"Personalization","score":0.5987441539764404},{"id":"https://openalex.org/keywords/word-error-rate","display_name":"Word error rate","score":0.592864990234375},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.5791865587234497},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.508120059967041},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.49097636342048645},{"id":"https://openalex.org/keywords/word","display_name":"Word (group theory)","score":0.4154689610004425},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.2947717607021332},{"id":"https://openalex.org/keywords/world-wide-web","display_name":"World Wide Web","score":0.07587367296218872}],"concepts":[{"id":"https://openalex.org/C147168706","wikidata":"https://www.wikidata.org/wiki/Q1457734","display_name":"Recurrent neural network","level":3,"score":0.8061233162879944},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8016386032104492},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.6416541934013367},{"id":"https://openalex.org/C183003079","wikidata":"https://www.wikidata.org/wiki/Q1000371","display_name":"Personalization","level":2,"score":0.5987441539764404},{"id":"https://openalex.org/C40969351","wikidata":"https://www.wikidata.org/wiki/Q3516228","display_name":"Word error rate","level":2,"score":0.592864990234375},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.5791865587234497},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.508120059967041},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.49097636342048645},{"id":"https://openalex.org/C90805587","wikidata":"https://www.wikidata.org/wiki/Q10944557","display_name":"Word (group theory)","level":2,"score":0.4154689610004425},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.2947717607021332},{"id":"https://openalex.org/C136764020","wikidata":"https://www.wikidata.org/wiki/Q466","display_name":"World Wide Web","level":1,"score":0.07587367296218872},{"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/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.0},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/slt48900.2021.9383560","is_oa":false,"landing_page_url":"https://doi.org/10.1109/slt48900.2021.9383560","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2021 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":0.6000000238418579}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":40,"referenced_works":["https://openalex.org/W648786980","https://openalex.org/W854541894","https://openalex.org/W1522301498","https://openalex.org/W1828163288","https://openalex.org/W1915251500","https://openalex.org/W2327501763","https://openalex.org/W2407080277","https://openalex.org/W2750499125","https://openalex.org/W2799800213","https://openalex.org/W2886319145","https://openalex.org/W2888779557","https://openalex.org/W2962760690","https://openalex.org/W2962784628","https://openalex.org/W2962824709","https://openalex.org/W2963240019","https://openalex.org/W2963250244","https://openalex.org/W2963382687","https://openalex.org/W2963827914","https://openalex.org/W2963979492","https://openalex.org/W2964121744","https://openalex.org/W2972625221","https://openalex.org/W2972630480","https://openalex.org/W2972880214","https://openalex.org/W2973172693","https://openalex.org/W3008480565","https://openalex.org/W3008525923","https://openalex.org/W3008898571","https://openalex.org/W3015315932","https://openalex.org/W3015486229","https://openalex.org/W3095311338","https://openalex.org/W3097794466","https://openalex.org/W3144557079","https://openalex.org/W6621543089","https://openalex.org/W6623517193","https://openalex.org/W6631190155","https://openalex.org/W6638749077","https://openalex.org/W6640059789","https://openalex.org/W6713762819","https://openalex.org/W6775968916","https://openalex.org/W6785229415"],"related_works":["https://openalex.org/W2160451571","https://openalex.org/W1566315437","https://openalex.org/W2151348424","https://openalex.org/W2594897229","https://openalex.org/W4290708361","https://openalex.org/W4221142855","https://openalex.org/W2050138804","https://openalex.org/W767271433","https://openalex.org/W2129812225","https://openalex.org/W1507733787"],"abstract_inverted_index":{"End-to-end":[0],"models":[1,41,59,105],"in":[2,10,16,22,62,117],"general,":[3],"and":[4,32,60,98,134,146,158],"Recurrent":[5],"Neural":[6],"Network":[7],"Transducer":[8],"(RNN-T)":[9],"particular,":[11],"have":[12],"gained":[13],"significant":[14],"traction":[15],"the":[17,23,54,89,92,141,148],"automatic":[18],"speech":[19],"recognition":[20,160],"community":[21],"last":[24],"few":[25],"years":[26],"due":[27,52],"to":[28,45,48,53,77,81,125],"their":[29],"simplicity,":[30],"compactness,":[31],"excellent":[33],"performance":[34],"on":[35,153],"generic":[36],"transcription":[37],"tasks.":[38],"However,":[39],"these":[40,113],"are":[42,161],"more":[43,107],"challenging":[44],"personalize":[46],"compared":[47,124],"traditional":[49],"hybrid":[50,151],"systems":[51,152],"lack":[55],"of":[56,94,143],"external":[57],"language":[58,104],"difficulties":[61],"recognizing":[63],"rare":[64,83],"long-tail":[65],"words,":[66],"specifically":[67],"entity":[68,159],"names.":[69],"In":[70],"this":[71],"work,":[72],"we":[73],"present":[74],"novel":[75],"techniques":[76,115],"improve":[78],"RNN-T's":[79],"ability":[80],"model":[82],"WordPieces,":[84],"infuse":[85],"extra":[86],"information":[87],"into":[88],"encoder,":[90],"enable":[91],"use":[93,154],"alternative":[95],"graphemic":[96],"pronunciations,":[97],"perform":[99],"deep":[100],"fusion":[101,133],"with":[102,150],"personalized":[103],"for":[106],"robust":[108],"biasing.":[109],"We":[110],"show":[111],"that":[112],"combined":[114],"result":[116],"15.4%-34.5%":[118],"relative":[119],"Word":[120],"Error":[121],"Rate":[122],"improvement":[123],"a":[126],"strong":[127],"RNN-T":[128,144],"baseline":[129],"which":[130],"uses":[131],"shallow":[132],"text-to-speech":[135],"augmentation.":[136],"Our":[137],"work":[138],"helps":[139],"push":[140],"boundary":[142],"personalization":[145],"close":[147],"gap":[149],"cases":[155],"where":[156],"biasing":[157],"crucial.":[162]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":11},{"year":2024,"cited_by_count":9},{"year":2023,"cited_by_count":20},{"year":2022,"cited_by_count":13},{"year":2021,"cited_by_count":10}],"updated_date":"2026-06-02T09:04:35.204637","created_date":"2025-10-10T00:00:00"}