{"id":"https://openalex.org/W3197037302","doi":"https://doi.org/10.21437/interspeech.2021-586","title":"HMM-Free Encoder Pre-Training for Streaming RNN Transducer","display_name":"HMM-Free Encoder Pre-Training for Streaming RNN Transducer","publication_year":2021,"publication_date":"2021-08-27","ids":{"openalex":"https://openalex.org/W3197037302","doi":"https://doi.org/10.21437/interspeech.2021-586","mag":"3197037302"},"language":"en","primary_location":{"id":"doi:10.21437/interspeech.2021-586","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2021-586","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Interspeech 2021","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/A5101623835","display_name":"Lu Huang","orcid":"https://orcid.org/0000-0002-1588-8669"},"institutions":[{"id":"https://openalex.org/I4210126929","display_name":"CE Technologies (United Kingdom)","ror":"https://ror.org/02pw67e25","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210126929"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Lu Huang","raw_affiliation_strings":["Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China","institution_ids":["https://openalex.org/I4210126929"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113877680","display_name":"Jingyu Sun","orcid":null},"institutions":[{"id":"https://openalex.org/I4210126929","display_name":"CE Technologies (United Kingdom)","ror":"https://ror.org/02pw67e25","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210126929"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Jingyu Sun","raw_affiliation_strings":["Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China","institution_ids":["https://openalex.org/I4210126929"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5031939484","display_name":"Yufeng Tang","orcid":null},"institutions":[{"id":"https://openalex.org/I4210126929","display_name":"CE Technologies (United Kingdom)","ror":"https://ror.org/02pw67e25","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210126929"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Yufeng Tang","raw_affiliation_strings":["Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China","institution_ids":["https://openalex.org/I4210126929"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101505189","display_name":"Junfeng Hou","orcid":"https://orcid.org/0000-0003-3635-5332"},"institutions":[{"id":"https://openalex.org/I4210126929","display_name":"CE Technologies (United Kingdom)","ror":"https://ror.org/02pw67e25","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210126929"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Junfeng Hou","raw_affiliation_strings":["Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China","institution_ids":["https://openalex.org/I4210126929"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103181178","display_name":"Jinkun Chen","orcid":"https://orcid.org/0000-0003-4657-1012"},"institutions":[{"id":"https://openalex.org/I4210126929","display_name":"CE Technologies (United Kingdom)","ror":"https://ror.org/02pw67e25","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210126929"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Jinkun Chen","raw_affiliation_strings":["Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China","institution_ids":["https://openalex.org/I4210126929"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100433129","display_name":"Jun Zhang","orcid":"https://orcid.org/0000-0002-2510-7205"},"institutions":[{"id":"https://openalex.org/I4210126929","display_name":"CE Technologies (United Kingdom)","ror":"https://ror.org/02pw67e25","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210126929"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Jun Zhang","raw_affiliation_strings":["Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China","institution_ids":["https://openalex.org/I4210126929"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5110851569","display_name":"Zejun Ma","orcid":null},"institutions":[{"id":"https://openalex.org/I4210126929","display_name":"CE Technologies (United Kingdom)","ror":"https://ror.org/02pw67e25","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210126929"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Zejun Ma","raw_affiliation_strings":["Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Speech & Audio Team, AI Lab, ByteDance Inc., Beijing, China","institution_ids":["https://openalex.org/I4210126929"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I4210126929"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1797","last_page":"1801"},"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9995999932289124,"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":0.9958000183105469,"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/initialization","display_name":"Initialization","score":0.8537572622299194},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.850531816482544},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.8027931451797485},{"id":"https://openalex.org/keywords/recurrent-neural-network","display_name":"Recurrent neural network","score":0.7933791875839233},{"id":"https://openalex.org/keywords/frame","display_name":"Frame (networking)","score":0.6517192721366882},{"id":"https://openalex.org/keywords/hidden-markov-model","display_name":"Hidden Markov model","score":0.6010695695877075},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5948468446731567},{"id":"https://openalex.org/keywords/lexicon","display_name":"Lexicon","score":0.5570926666259766},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.4856332838535309},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4781445264816284},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.38667571544647217}],"concepts":[{"id":"https://openalex.org/C114466953","wikidata":"https://www.wikidata.org/wiki/Q6034165","display_name":"Initialization","level":2,"score":0.8537572622299194},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.850531816482544},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.8027931451797485},{"id":"https://openalex.org/C147168706","wikidata":"https://www.wikidata.org/wiki/Q1457734","display_name":"Recurrent neural network","level":3,"score":0.7933791875839233},{"id":"https://openalex.org/C126042441","wikidata":"https://www.wikidata.org/wiki/Q1324888","display_name":"Frame (networking)","level":2,"score":0.6517192721366882},{"id":"https://openalex.org/C23224414","wikidata":"https://www.wikidata.org/wiki/Q176769","display_name":"Hidden Markov model","level":2,"score":0.6010695695877075},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5948468446731567},{"id":"https://openalex.org/C2778121359","wikidata":"https://www.wikidata.org/wiki/Q8096","display_name":"Lexicon","level":2,"score":0.5570926666259766},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4856332838535309},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4781445264816284},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.38667571544647217},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.0},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.21437/interspeech.2021-586","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2021-586","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Interspeech 2021","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Quality Education","score":0.7699999809265137,"id":"https://metadata.un.org/sdg/4"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":34,"referenced_works":["https://openalex.org/W1494198834","https://openalex.org/W1524333225","https://openalex.org/W1828163288","https://openalex.org/W2102113734","https://openalex.org/W2143612262","https://openalex.org/W2293634267","https://openalex.org/W2327501763","https://openalex.org/W2597757402","https://openalex.org/W2766219058","https://openalex.org/W2799800213","https://openalex.org/W2888884590","https://openalex.org/W2913178639","https://openalex.org/W2935756939","https://openalex.org/W2936774411","https://openalex.org/W2936993002","https://openalex.org/W2962760690","https://openalex.org/W2962826786","https://openalex.org/W2963414781","https://openalex.org/W2963827914","https://openalex.org/W2964084166","https://openalex.org/W2973040747","https://openalex.org/W3011207290","https://openalex.org/W3015194534","https://openalex.org/W3015315932","https://openalex.org/W3015686596","https://openalex.org/W3015927303","https://openalex.org/W3016010032","https://openalex.org/W3095410713","https://openalex.org/W3096518646","https://openalex.org/W3096702180","https://openalex.org/W3162649911","https://openalex.org/W3163152557","https://openalex.org/W4295312788","https://openalex.org/W4385245566"],"related_works":["https://openalex.org/W3204184292","https://openalex.org/W3176564347","https://openalex.org/W1985458517","https://openalex.org/W2355833770","https://openalex.org/W3031039437","https://openalex.org/W2805949325","https://openalex.org/W4300536205","https://openalex.org/W2160451571","https://openalex.org/W2495256954","https://openalex.org/W2259317772"],"abstract_inverted_index":{"This":[0],"work":[1,113],"describes":[2],"an":[3,77],"encoder":[4,82],"pre-training":[5,40,83,134],"procedure":[6],"using":[7,73,119],"frame-wise":[8,48,62,117],"label":[9,118],"to":[10,35,65,95,114,161],"improve":[11],"the":[12,51,90,111,132,141,147,153],"training":[13],"of":[14,42,92],"streaming":[15,67],"recurrent":[16],"neural":[17],"network":[18],"transducer":[19],"(RNN-T)":[20],"model.Streaming":[21],"RNN-T":[22,43],"trained":[23],"from":[24],"scratch":[25],"usually":[26],"performs":[27],"worse":[28],"than":[29],"nonstreaming":[30],"RNN-T.Although":[31],"it":[32,158],"is":[33,53,70,84,86,110,155,159],"common":[34],"address":[36],"this":[37,60,109],"issue":[38],"through":[39],"components":[41],"with":[44,137],"other":[45],"criteria":[46],"or":[47],"alignment":[49,52],"guidance,":[50],"not":[54],"easily":[55],"available":[56],"in":[57],"end-to-end":[58],"manner.In":[59],"work,":[61],"alignment,":[63],"used":[64],"pre-train":[66],"RNN-T's":[68],"encoder,":[69],"generated":[71],"without":[72,164],"a":[74],"HMM-based":[75,116],"system.Therefore":[76],"allneural":[78],"framework":[79],"equipping":[80],"HMM-free":[81],"constructed.This":[85],"achieved":[87],"by":[88,143,150],"expanding":[89,102],"spikes":[91],"CTC":[93,120],"model":[94,121],"their":[96],"left/right":[97],"blank":[98],"frames,":[99],"and":[100,127,146],"two":[101],"strategies":[103],"are":[104],"proposed.To":[105],"our":[106],"best":[107],"knowledge,":[108],"first":[112],"simulate":[115],"for":[122],"pre-training.Experiments":[123],"conducted":[124],"on":[125],"LibriSpeech":[126],"MLS":[128],"English":[129],"tasks":[130],"show":[131],"proposed":[133],"procedure,":[135],"compared":[136],"random":[138],"initialization,":[139],"reduces":[140],"WER":[142],"relatively":[144],"5%\u223c11%":[145],"emission":[148],"latency":[149],"60":[151],"ms.Besides,":[152],"method":[154],"lexicon-free,":[156],"so":[157],"friendly":[160],"new":[162],"languages":[163],"manually":[165],"designed":[166],"lexicon.":[167]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}