{"id":"https://openalex.org/W3161425572","doi":"https://doi.org/10.1109/icassp39728.2021.9414339","title":"Improving RNN Transducer Modeling for Small-Footprint Keyword Spotting","display_name":"Improving RNN Transducer Modeling for Small-Footprint Keyword Spotting","publication_year":2021,"publication_date":"2021-05-13","ids":{"openalex":"https://openalex.org/W3161425572","doi":"https://doi.org/10.1109/icassp39728.2021.9414339","mag":"3161425572"},"language":"en","primary_location":{"id":"doi:10.1109/icassp39728.2021.9414339","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp39728.2021.9414339","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","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/A5080127413","display_name":"Yao Tian","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Yao Tian","raw_affiliation_strings":["Bytedance AI Lab"],"affiliations":[{"raw_affiliation_string":"Bytedance AI Lab","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5061055212","display_name":"Haitao Yao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Haitao Yao","raw_affiliation_strings":["Bytedance AI Lab"],"affiliations":[{"raw_affiliation_string":"Bytedance AI Lab","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5052057753","display_name":"Meng Cai","orcid":"https://orcid.org/0000-0002-0711-5949"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Meng Cai","raw_affiliation_strings":["Bytedance AI Lab"],"affiliations":[{"raw_affiliation_string":"Bytedance AI Lab","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101553467","display_name":"Yaming Liu","orcid":"https://orcid.org/0000-0002-3982-001X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yaming Liu","raw_affiliation_strings":["Bytedance AI Lab"],"affiliations":[{"raw_affiliation_string":"Bytedance AI Lab","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5110851569","display_name":"Zejun Ma","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zejun Ma","raw_affiliation_strings":["Bytedance AI Lab"],"affiliations":[{"raw_affiliation_string":"Bytedance AI Lab","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5080127413"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":2.1122,"has_fulltext":false,"cited_by_count":22,"citation_normalized_percentile":{"value":0.892333,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"5624","last_page":"5628"},"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/T10860","display_name":"Speech and Audio Processing","score":0.9977999925613403,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9896000027656555,"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/keyword-spotting","display_name":"Keyword spotting","score":0.9430636167526245},{"id":"https://openalex.org/keywords/recurrent-neural-network","display_name":"Recurrent neural network","score":0.9178248643875122},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8394663333892822},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.5950219631195068},{"id":"https://openalex.org/keywords/cross-entropy","display_name":"Cross entropy","score":0.5711715221405029},{"id":"https://openalex.org/keywords/spotting","display_name":"Spotting","score":0.5262077450752258},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5151247382164001},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5147745609283447},{"id":"https://openalex.org/keywords/data-modeling","display_name":"Data modeling","score":0.48988789319992065},{"id":"https://openalex.org/keywords/connectionism","display_name":"Connectionism","score":0.4816243052482605},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.457175076007843},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.45382747054100037},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.4530957043170929},{"id":"https://openalex.org/keywords/memory-footprint","display_name":"Memory footprint","score":0.4452108144760132},{"id":"https://openalex.org/keywords/footprint","display_name":"Footprint","score":0.42322731018066406},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.34078872203826904},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.3221104145050049},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.13164561986923218},{"id":"https://openalex.org/keywords/database","display_name":"Database","score":0.10079368948936462}],"concepts":[{"id":"https://openalex.org/C2781213101","wikidata":"https://www.wikidata.org/wiki/Q6398558","display_name":"Keyword spotting","level":2,"score":0.9430636167526245},{"id":"https://openalex.org/C147168706","wikidata":"https://www.wikidata.org/wiki/Q1457734","display_name":"Recurrent neural network","level":3,"score":0.9178248643875122},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8394663333892822},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.5950219631195068},{"id":"https://openalex.org/C167981619","wikidata":"https://www.wikidata.org/wiki/Q1685498","display_name":"Cross entropy","level":3,"score":0.5711715221405029},{"id":"https://openalex.org/C2779506182","wikidata":"https://www.wikidata.org/wiki/Q7580141","display_name":"Spotting","level":2,"score":0.5262077450752258},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5151247382164001},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5147745609283447},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.48988789319992065},{"id":"https://openalex.org/C8521452","wikidata":"https://www.wikidata.org/wiki/Q203790","display_name":"Connectionism","level":3,"score":0.4816243052482605},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.457175076007843},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.45382747054100037},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.4530957043170929},{"id":"https://openalex.org/C74912251","wikidata":"https://www.wikidata.org/wiki/Q6815727","display_name":"Memory footprint","level":2,"score":0.4452108144760132},{"id":"https://openalex.org/C132943942","wikidata":"https://www.wikidata.org/wiki/Q2562511","display_name":"Footprint","level":2,"score":0.42322731018066406},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.34078872203826904},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3221104145050049},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.13164561986923218},{"id":"https://openalex.org/C77088390","wikidata":"https://www.wikidata.org/wiki/Q8513","display_name":"Database","level":1,"score":0.10079368948936462},{"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/C151730666","wikidata":"https://www.wikidata.org/wiki/Q7205","display_name":"Paleontology","level":1,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp39728.2021.9414339","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp39728.2021.9414339","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2021 - 2021 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":20,"referenced_works":["https://openalex.org/W1828163288","https://openalex.org/W2034940213","https://openalex.org/W2327501763","https://openalex.org/W2407023693","https://openalex.org/W2507319753","https://openalex.org/W2797759721","https://openalex.org/W2962707338","https://openalex.org/W2962760690","https://openalex.org/W2963308316","https://openalex.org/W2963414149","https://openalex.org/W2963970535","https://openalex.org/W2963977978","https://openalex.org/W3011207290","https://openalex.org/W3015315932","https://openalex.org/W3015639015","https://openalex.org/W3015686596","https://openalex.org/W3048653539","https://openalex.org/W3095184753","https://openalex.org/W6638749077","https://openalex.org/W6714171909"],"related_works":["https://openalex.org/W4367000378","https://openalex.org/W2889176648","https://openalex.org/W2964187693","https://openalex.org/W2766772137","https://openalex.org/W3007438732","https://openalex.org/W4287864706","https://openalex.org/W3161425572","https://openalex.org/W3000294609","https://openalex.org/W4289285304","https://openalex.org/W2963628261"],"abstract_inverted_index":{"The":[0,96],"recurrent":[1],"neural":[2,132],"network":[3,31,57,133,140,146],"transducer":[4],"(RNN-T)":[5],"model":[6,35,38,158,164],"has":[7,59],"been":[8],"proved":[9],"effective":[10],"for":[11,70,138,151],"keyword":[12,46,53,72],"spotting":[13,73],"(KWS)":[14],"recently.":[15],"However,":[16],"compared":[17],"with":[18,101,114,124,159,166,195],"cross-entropy":[19],"(CE)":[20],"or":[21],"connectionist":[22],"temporal":[23],"classification":[24],"(CTC)":[25],"based":[26],"models,":[27],"the":[28,33,37,45,52,55,67,80,94,109,120,130,136,157,177,186,192],"additional":[29],"prediction":[30,56,139,145],"in":[32,74],"RNN-T":[34,68,110,188],"increases":[36],"size":[39],"and":[40,105],"computational":[41],"cost.":[42],"Besides,":[43],"since":[44],"training":[47,86],"data":[48,104,116,172],"usually":[49],"only":[50,119],"contain":[51],"sequence,":[54],"might":[58],"over-fitting":[60,81],"problems.":[61],"In":[62],"this":[63],"paper,":[64],"we":[65,83,128,154],"improve":[66,156],"modeling":[69],"small-footprint":[71],"three":[75],"aspects.":[76],"First,":[77],"to":[78,93,134,175],"address":[79],"issue,":[82],"explore":[84],"multi-task":[85],"where":[87,162],"a":[88,163,181,196],"CTC":[89,97],"loss":[90,98,111],"is":[91,99,112,122,173],"added":[92],"encoder.":[95],"calculated":[100,113],"both":[102],"KWS":[103,125,178],"ASR":[106,115,171],"data,":[107],"while":[108],"so":[117],"that":[118],"encoder":[121],"augmented":[123],"data.":[126],"Second,":[127],"use":[129],"feed-forward":[131],"replace":[135],"LSTM":[137],"modeling.":[141],"Thus":[142],"all":[143],"possible":[144],"outputs":[147],"could":[148],"be":[149],"pre-computed":[150],"decoding.":[152],"Third,":[153],"further":[155],"transfer":[160],"learning,":[161],"trained":[165],"160":[167],"thousand":[168],"hours":[169],"of":[170],"used":[174],"initialize":[176],"model.":[179],"On":[180],"self-collected":[182],"far-field":[183],"wake-word":[184],"testset,":[185],"proposed":[187],"system":[189],"greatly":[190],"improves":[191],"performance":[193],"comparing":[194],"strong":[197],"\"keyword-filler\"":[198],"baseline.":[199]},"counts_by_year":[{"year":2025,"cited_by_count":7},{"year":2024,"cited_by_count":5},{"year":2023,"cited_by_count":4},{"year":2022,"cited_by_count":4},{"year":2021,"cited_by_count":2}],"updated_date":"2026-04-03T22:45:19.894376","created_date":"2025-10-10T00:00:00"}