{"id":"https://openalex.org/W3198368663","doi":"https://doi.org/10.21437/interspeech.2021-1272","title":"Dynamic Encoder Transducer: A Flexible Solution for Trading Off Accuracy for Latency","display_name":"Dynamic Encoder Transducer: A Flexible Solution for Trading Off Accuracy for Latency","publication_year":2021,"publication_date":"2021-08-27","ids":{"openalex":"https://openalex.org/W3198368663","doi":"https://doi.org/10.21437/interspeech.2021-1272","mag":"3198368663"},"language":"en","primary_location":{"id":"doi:10.21437/interspeech.2021-1272","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2021-1272","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":"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/A5103247973","display_name":"Yangyang Shi","orcid":"https://orcid.org/0000-0001-5297-4155"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Yangyang Shi","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034208747","display_name":"Varun Nagaraja","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Varun Nagaraja","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101579713","display_name":"Chunyang Wu","orcid":"https://orcid.org/0000-0002-0269-3555"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chunyang Wu","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5074237839","display_name":"Jay Mahadeokar","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jay Mahadeokar","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103098852","display_name":"Duc Le","orcid":"https://orcid.org/0000-0001-9490-2563"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Duc Le","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5032640894","display_name":"Rohit Prabhavalkar","orcid":"https://orcid.org/0000-0001-5331-6058"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rohit Prabhavalkar","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5056584611","display_name":"Alex Xiao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Alex Xiao","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5087421101","display_name":"Ching-Feng Yeh","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ching-Feng Yeh","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5051075420","display_name":"Julian Chan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Julian Chan","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5047073253","display_name":"Christian Fuegen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Christian Fuegen","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066166549","display_name":"Ozlem Kalinli","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ozlem Kalinli","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5041313589","display_name":"Michael L. Seltzer","orcid":"https://orcid.org/0000-0003-3474-2451"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Michael L. Seltzer","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":12,"corresponding_author_ids":["https://openalex.org/A5103247973"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.0878,"has_fulltext":false,"cited_by_count":9,"citation_normalized_percentile":{"value":0.81943962,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"2042","last_page":"2046"},"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/T10860","display_name":"Speech and Audio Processing","score":0.9990000128746033,"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/T11309","display_name":"Music and Audio Processing","score":0.9948999881744385,"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/encoder","display_name":"Encoder","score":0.7928006649017334},{"id":"https://openalex.org/keywords/transducer","display_name":"Transducer","score":0.7432371377944946},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7257680296897888},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.7205494046211243},{"id":"https://openalex.org/keywords/real-time-computing","display_name":"Real-time computing","score":0.3998800814151764},{"id":"https://openalex.org/keywords/telecommunications","display_name":"Telecommunications","score":0.14098161458969116},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.12220251560211182},{"id":"https://openalex.org/keywords/electrical-engineering","display_name":"Electrical engineering","score":0.11146271228790283},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.08224168419837952}],"concepts":[{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.7928006649017334},{"id":"https://openalex.org/C56318395","wikidata":"https://www.wikidata.org/wiki/Q215928","display_name":"Transducer","level":2,"score":0.7432371377944946},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7257680296897888},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.7205494046211243},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.3998800814151764},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.14098161458969116},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.12220251560211182},{"id":"https://openalex.org/C119599485","wikidata":"https://www.wikidata.org/wiki/Q43035","display_name":"Electrical engineering","level":1,"score":0.11146271228790283},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.08224168419837952}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.21437/interspeech.2021-1272","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2021-1272","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":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2072454424","https://openalex.org/W4384820447","https://openalex.org/W2185942010","https://openalex.org/W2260725127","https://openalex.org/W2117438306","https://openalex.org/W1992056405","https://openalex.org/W767846903","https://openalex.org/W2954284861"],"abstract_inverted_index":{"We":[0,43],"propose":[1],"a":[2,102,161,168,175,186],"dynamic":[3],"encoder":[4,65,115,134,177,188],"transducer":[5],"(DET)":[6],"for":[7,54,178,189],"on-device":[8],"speech":[9],"recognition.":[10],"One":[11],"DET":[12,32,55,100,117,136,156],"model":[13,143,163],"scales":[14],"to":[15,36],"multiple":[16,82],"devices":[17],"with":[18,84,138,164],"different":[19,34,38,85],"computation":[20],"capacities":[21],"without":[22],"retraining":[23],"or":[24],"finetuning.":[25],"To":[26],"trading":[27],"off":[28],"accuracy":[29,104,159],"and":[30,45,50,95,105,185],"latency,":[31],"assigns":[33],"encoders":[35,83],"decode":[37],"parts":[39],"of":[40,124,182],"an":[41],"utterance.":[42],"apply":[44],"compare":[46],"the":[47,51,68,113,120,125,142,153,179,190],"layer":[48,58,74],"dropout":[49,59,75],"collaborative":[52,139],"learning":[53,79,140],"training.":[56],"The":[57,132],"method":[60],"that":[61,99,112],"randomly":[62],"drops":[63],"out":[64],"layers":[66],"in":[67,76,87,116,135],"training":[69],"phase,":[70],"can":[71],"do":[72],"on-demand":[73],"decoding.":[77],"Collaborative":[78],"jointly":[80],"trains":[81],"depths":[86],"one":[88,183],"single":[89],"model.":[90],"Experiment":[91],"results":[92],"on":[93,109,167],"Librispeech":[94,110],"in-house":[96,170],"data":[97,171],"show":[98,111],"provides":[101],"flexible":[103],"latency":[106,166],"trade-off.":[107],"Results":[108],"full-size":[114,154,187],"relatively":[118],"reduces":[119,141],"word":[121],"error":[122],"rate":[123],"same":[126],"size":[127,144],"baseline":[128,162],"by":[129,145,173],"over":[130],"8%.":[131],"lightweight":[133,176],"trained":[137],"25%":[146],"but":[147],"still":[148],"gets":[149,157],"similar":[150,158],"WER":[151],"as":[152,160],"baseline.":[155],"better":[165],"large":[169],"set":[172],"assigning":[174],"beginning":[180],"part":[181],"utterance":[184],"rest.":[191]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2024,"cited_by_count":1},{"year":2023,"cited_by_count":5},{"year":2022,"cited_by_count":2}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}