{"id":"https://openalex.org/W3016155933","doi":"https://doi.org/10.1109/icassp40776.2020.9054387","title":"High-Accuracy and Low-Latency Speech Recognition with Two-Head Contextual Layer Trajectory LSTM Model","display_name":"High-Accuracy and Low-Latency Speech Recognition with Two-Head Contextual Layer Trajectory LSTM Model","publication_year":2020,"publication_date":"2020-04-09","ids":{"openalex":"https://openalex.org/W3016155933","doi":"https://doi.org/10.1109/icassp40776.2020.9054387","mag":"3016155933"},"language":"en","primary_location":{"id":"doi:10.1109/icassp40776.2020.9054387","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp40776.2020.9054387","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2020 - 2020 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/A5100365053","display_name":"Jinyu Li","orcid":"https://orcid.org/0000-0002-1089-9748"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jinyu Li","raw_affiliation_strings":["Microsoft Speech and Language Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Speech and Language Group","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101987526","display_name":"Rui Zhao","orcid":"https://orcid.org/0000-0002-9699-9984"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Rui Zhao","raw_affiliation_strings":["Microsoft Speech and Language Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Speech and Language Group","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5090360853","display_name":"Eric Sun","orcid":"https://orcid.org/0000-0001-8805-9864"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Eric Sun","raw_affiliation_strings":["Microsoft Speech and Language Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Speech and Language Group","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5048201022","display_name":"Jeremy H. M. Wong","orcid":"https://orcid.org/0000-0003-3742-7510"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jeremy H. M. Wong","raw_affiliation_strings":["Microsoft Speech and Language Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Speech and Language Group","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101814582","display_name":"Amit Das","orcid":"https://orcid.org/0000-0001-7733-2976"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Amit Das","raw_affiliation_strings":["Microsoft Speech and Language Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Speech and Language Group","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101749753","display_name":"Zhong Meng","orcid":"https://orcid.org/0000-0001-7814-5929"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhong Meng","raw_affiliation_strings":["Microsoft Speech and Language Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Speech and Language Group","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5077401426","display_name":"Yifan Gong","orcid":"https://orcid.org/0000-0001-8786-3391"},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yifan Gong","raw_affiliation_strings":["Microsoft Speech and Language Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Microsoft Speech and Language Group","institution_ids":["https://openalex.org/I1290206253"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I1290206253"],"apc_list":null,"apc_paid":null,"fwci":2.6449,"has_fulltext":false,"cited_by_count":24,"citation_normalized_percentile":{"value":0.92036344,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":90,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"7699","last_page":"7703"},"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":0.9991999864578247,"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":0.9980999827384949,"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/computer-science","display_name":"Computer science","score":0.8557099103927612},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.7142997980117798},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.7119940519332886},{"id":"https://openalex.org/keywords/discriminative-model","display_name":"Discriminative model","score":0.6593233346939087},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.521872878074646},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.44857582449913025},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.324896901845932}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8557099103927612},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.7142997980117798},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.7119940519332886},{"id":"https://openalex.org/C97931131","wikidata":"https://www.wikidata.org/wiki/Q5282087","display_name":"Discriminative model","level":2,"score":0.6593233346939087},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.521872878074646},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.44857582449913025},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.324896901845932},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"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/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp40776.2020.9054387","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp40776.2020.9054387","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Reduced inequalities","id":"https://metadata.un.org/sdg/10","score":0.6200000047683716}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":46,"referenced_works":["https://openalex.org/W38527073","https://openalex.org/W1771459135","https://openalex.org/W1821462560","https://openalex.org/W2003123121","https://openalex.org/W2064675550","https://openalex.org/W2291522532","https://openalex.org/W2293634267","https://openalex.org/W2396144419","https://openalex.org/W2396384435","https://openalex.org/W2402040300","https://openalex.org/W2475988411","https://openalex.org/W2507699225","https://openalex.org/W2513938599","https://openalex.org/W2515439472","https://openalex.org/W2545177271","https://openalex.org/W2586630418","https://openalex.org/W2735006420","https://openalex.org/W2787290364","https://openalex.org/W2911629330","https://openalex.org/W2936123380","https://openalex.org/W2939248538","https://openalex.org/W2962760690","https://openalex.org/W2962765220","https://openalex.org/W2962824709","https://openalex.org/W2962965465","https://openalex.org/W2963211739","https://openalex.org/W2963326356","https://openalex.org/W2963738441","https://openalex.org/W2963827914","https://openalex.org/W2963917928","https://openalex.org/W2964084166","https://openalex.org/W2964272710","https://openalex.org/W2972389417","https://openalex.org/W2973122799","https://openalex.org/W2992632249","https://openalex.org/W3007227084","https://openalex.org/W3095189764","https://openalex.org/W4235542672","https://openalex.org/W4294619417","https://openalex.org/W6638205174","https://openalex.org/W6638523607","https://openalex.org/W6712847557","https://openalex.org/W6728910023","https://openalex.org/W6747158283","https://openalex.org/W6748052321","https://openalex.org/W6770812743"],"related_works":["https://openalex.org/W2965546495","https://openalex.org/W4389116644","https://openalex.org/W2153315159","https://openalex.org/W3103844505","https://openalex.org/W259157601","https://openalex.org/W4205463238","https://openalex.org/W2761785940","https://openalex.org/W2110523656","https://openalex.org/W2404514746","https://openalex.org/W1652783584"],"abstract_inverted_index":{"While":[0],"the":[1,80,103,127,158,164,175],"community":[2],"keeps":[3],"promoting":[4],"end-to-end":[5],"models":[6,18,38,58],"over":[7,174],"conventional":[8,36,54,176],"hybrid":[9,37,55],"models,":[10],"which":[11,78,120],"usually":[12],"are":[13],"long":[14],"short-term":[15],"memory":[16],"(LSTM)":[17],"trained":[19,139],"with":[20,106,140,151,154,163,180],"a":[21,27,72,116,131,169,181],"cross":[22],"entropy":[23],"criterion":[24],"followed":[25],"by":[26],"sequence":[28],"discriminative":[29],"training":[30,104,147,166],"criterion,":[31],"we":[32,47,70,114],"argue":[33],"that":[34],"such":[35],"can":[39],"still":[40],"be":[41],"significantly":[42],"improved.":[43],"In":[44],"this":[45],"paper,":[46],"detail":[48],"our":[49],"recent":[50],"efforts":[51],"to":[52,92,135],"improve":[53,102],"LSTM":[56,76,177],"acoustic":[57,98,178],"for":[59,96],"high-accuracy":[60],"and":[61,83,87,126,149],"low-latency":[62],"automatic":[63],"speech":[64],"recognition.":[65],"To":[66,110],"achieve":[67],"high":[68],"accuracy,":[69],"use":[71],"contextual":[73],"layer":[74],"trajectory":[75],"(cltLSTM),":[77],"decouples":[79],"temporal":[81],"modeling":[82],"target":[84],"classification":[85],"tasks,":[86],"incorporates":[88],"future":[89],"context":[90],"frames":[91],"get":[93],"more":[94],"information":[95],"accurate":[97],"modeling.":[99],"We":[100],"further":[101],"strategy":[105,167],"sequence-level":[107],"teacher-student":[108],"learning.":[109],"obtain":[111],"low":[112],"latency,":[113,133],"design":[115],"two-head":[117,160],"cltLSTM,":[118],"in":[119],"one":[121],"head":[122,129],"has":[123,130],"zero":[124],"latency":[125],"other":[128],"small":[132],"compared":[134],"an":[136],"LSTM.":[137],"When":[138],"Microsoft's":[141],"65":[142],"thousand":[143],"hours":[144],"of":[145],"anonymized":[146],"data":[148],"evaluated":[150],"test":[152],"sets":[153],"1.8":[155],"million":[156],"words,":[157],"proposed":[159,165],"cltLSTM":[161],"model":[162],"yields":[168],"28.2%":[170],"relative":[171],"WER":[172],"reduction":[173],"model,":[179],"similar":[182],"perceived":[183],"latency.":[184]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2023,"cited_by_count":1},{"year":2022,"cited_by_count":3},{"year":2021,"cited_by_count":15},{"year":2020,"cited_by_count":4}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}