{"id":"https://openalex.org/W3097246512","doi":"https://doi.org/10.1109/icassp39728.2021.9414797","title":"Optimize What Matters: Training DNN-Hmm Keyword Spotting Model Using End Metric","display_name":"Optimize What Matters: Training DNN-Hmm Keyword Spotting Model Using End Metric","publication_year":2021,"publication_date":"2021-05-13","ids":{"openalex":"https://openalex.org/W3097246512","doi":"https://doi.org/10.1109/icassp39728.2021.9414797","mag":"3097246512"},"language":"en","primary_location":{"id":"doi:10.1109/icassp39728.2021.9414797","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp39728.2021.9414797","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":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2011.01151","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5002740493","display_name":"Ashish Shrivastava","orcid":"https://orcid.org/0000-0002-2465-0708"},"institutions":[{"id":"https://openalex.org/I4210107260","display_name":"Apple (United Kingdom)","ror":"https://ror.org/01vpeym60","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210107260"]},{"id":"https://openalex.org/I4210141230","display_name":"Apple (Germany)","ror":"https://ror.org/03fjbw519","country_code":"DE","type":"company","lineage":["https://openalex.org/I4210141230","https://openalex.org/I4210153776"]}],"countries":["DE","GB"],"is_corresponding":false,"raw_author_name":"Ashish Shrivastava","raw_affiliation_strings":["Apple","[Apple]"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Apple","institution_ids":["https://openalex.org/I4210107260"]},{"raw_affiliation_string":"[Apple]","institution_ids":["https://openalex.org/I4210141230"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034953394","display_name":"Arnav Kundu","orcid":"https://orcid.org/0000-0003-2921-4312"},"institutions":[{"id":"https://openalex.org/I4210107260","display_name":"Apple (United Kingdom)","ror":"https://ror.org/01vpeym60","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210107260"]},{"id":"https://openalex.org/I4210141230","display_name":"Apple (Germany)","ror":"https://ror.org/03fjbw519","country_code":"DE","type":"company","lineage":["https://openalex.org/I4210141230","https://openalex.org/I4210153776"]}],"countries":["DE","GB"],"is_corresponding":false,"raw_author_name":"Arnav Kundu","raw_affiliation_strings":["Apple","[Apple]"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Apple","institution_ids":["https://openalex.org/I4210107260"]},{"raw_affiliation_string":"[Apple]","institution_ids":["https://openalex.org/I4210141230"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5110720542","display_name":"Chandra Dhir","orcid":null},"institutions":[{"id":"https://openalex.org/I4210107260","display_name":"Apple (United Kingdom)","ror":"https://ror.org/01vpeym60","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210107260"]},{"id":"https://openalex.org/I4210141230","display_name":"Apple (Germany)","ror":"https://ror.org/03fjbw519","country_code":"DE","type":"company","lineage":["https://openalex.org/I4210141230","https://openalex.org/I4210153776"]}],"countries":["DE","GB"],"is_corresponding":false,"raw_author_name":"Chandra Dhir","raw_affiliation_strings":["Apple","[Apple]"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Apple","institution_ids":["https://openalex.org/I4210107260"]},{"raw_affiliation_string":"[Apple]","institution_ids":["https://openalex.org/I4210141230"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113880378","display_name":"Devang Naik","orcid":"https://orcid.org/0009-0007-7838-1623"},"institutions":[{"id":"https://openalex.org/I4210107260","display_name":"Apple (United Kingdom)","ror":"https://ror.org/01vpeym60","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210107260"]},{"id":"https://openalex.org/I4210141230","display_name":"Apple (Germany)","ror":"https://ror.org/03fjbw519","country_code":"DE","type":"company","lineage":["https://openalex.org/I4210141230","https://openalex.org/I4210153776"]}],"countries":["DE","GB"],"is_corresponding":false,"raw_author_name":"Devang Naik","raw_affiliation_strings":["Apple","[Apple]"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Apple","institution_ids":["https://openalex.org/I4210107260"]},{"raw_affiliation_string":"[Apple]","institution_ids":["https://openalex.org/I4210141230"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5028613002","display_name":"Oncel Tuzel","orcid":null},"institutions":[{"id":"https://openalex.org/I4210107260","display_name":"Apple (United Kingdom)","ror":"https://ror.org/01vpeym60","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210107260"]},{"id":"https://openalex.org/I4210141230","display_name":"Apple (Germany)","ror":"https://ror.org/03fjbw519","country_code":"DE","type":"company","lineage":["https://openalex.org/I4210141230","https://openalex.org/I4210153776"]}],"countries":["DE","GB"],"is_corresponding":false,"raw_author_name":"Oncel Tuzel","raw_affiliation_strings":["Apple","[Apple]"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Apple","institution_ids":["https://openalex.org/I4210107260"]},{"raw_affiliation_string":"[Apple]","institution_ids":["https://openalex.org/I4210141230"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"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":"4000","last_page":"4004"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":0.9998999834060669,"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"}},"topics":[{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":0.9998999834060669,"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/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/T11309","display_name":"Music and Audio Processing","score":0.998199999332428,"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":0.8052752017974854},{"id":"https://openalex.org/keywords/hidden-markov-model","display_name":"Hidden Markov model","score":0.7908382415771484},{"id":"https://openalex.org/keywords/keyword-spotting","display_name":"Keyword spotting","score":0.7728912830352783},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.6639423966407776},{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.5433626174926758},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.48904237151145935},{"id":"https://openalex.org/keywords/spotting","display_name":"Spotting","score":0.4626188576221466},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.45556047558784485},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.44767817854881287},{"id":"https://openalex.org/keywords/performance-metric","display_name":"Performance metric","score":0.41823437809944153},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.3913664221763611}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8052752017974854},{"id":"https://openalex.org/C23224414","wikidata":"https://www.wikidata.org/wiki/Q176769","display_name":"Hidden Markov model","level":2,"score":0.7908382415771484},{"id":"https://openalex.org/C2781213101","wikidata":"https://www.wikidata.org/wiki/Q6398558","display_name":"Keyword spotting","level":2,"score":0.7728912830352783},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.6639423966407776},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.5433626174926758},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.48904237151145935},{"id":"https://openalex.org/C2779506182","wikidata":"https://www.wikidata.org/wiki/Q7580141","display_name":"Spotting","level":2,"score":0.4626188576221466},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.45556047558784485},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.44767817854881287},{"id":"https://openalex.org/C2780898871","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Performance metric","level":2,"score":0.41823437809944153},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3913664221763611},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C21547014","wikidata":"https://www.wikidata.org/wiki/Q1423657","display_name":"Operations management","level":1,"score":0.0},{"id":"https://openalex.org/C187736073","wikidata":"https://www.wikidata.org/wiki/Q2920921","display_name":"Management","level":1,"score":0.0}],"mesh":[],"locations_count":5,"locations":[{"id":"doi:10.1109/icassp39728.2021.9414797","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp39728.2021.9414797","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"},{"id":"pmh:oai:arXiv.org:2011.01151","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2011.01151","pdf_url":"https://arxiv.org/pdf/2011.01151","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"mag:3097246512","is_oa":true,"landing_page_url":"https://arxiv.org/pdf/2011.01151.pdf","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},{"id":"doi:10.48550/arxiv.2011.01151","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2011.01151","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"},{"id":"doi:10.17023/4gzh-8993","is_oa":true,"landing_page_url":"https://doi.org/10.17023/4gzh-8993","pdf_url":null,"source":{"id":"https://openalex.org/S7407051697","display_name":"IEEE RESOURCE CENTERS","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Audiovisual"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2011.01151","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2011.01151","pdf_url":"https://arxiv.org/pdf/2011.01151","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/16","score":0.6700000166893005,"display_name":"Peace, Justice and strong institutions"}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":false},"content_urls":{"pdf":"https://content.openalex.org/works/W3097246512.pdf"},"referenced_works_count":40,"referenced_works":["https://openalex.org/W1553469512","https://openalex.org/W1877570817","https://openalex.org/W2007857129","https://openalex.org/W2034365297","https://openalex.org/W2034940213","https://openalex.org/W2056986588","https://openalex.org/W2074671152","https://openalex.org/W2101346879","https://openalex.org/W2122797512","https://openalex.org/W2127141656","https://openalex.org/W2148154194","https://openalex.org/W2151244813","https://openalex.org/W2212465773","https://openalex.org/W2295088914","https://openalex.org/W2295405790","https://openalex.org/W2407023693","https://openalex.org/W2558258029","https://openalex.org/W2570343428","https://openalex.org/W2613718673","https://openalex.org/W2748659049","https://openalex.org/W2769912137","https://openalex.org/W2886949521","https://openalex.org/W2889511491","https://openalex.org/W2903910878","https://openalex.org/W2963037989","https://openalex.org/W2964121744","https://openalex.org/W2964187693","https://openalex.org/W2980286501","https://openalex.org/W3096831963","https://openalex.org/W6620707391","https://openalex.org/W6631190155","https://openalex.org/W6688089860","https://openalex.org/W6697212960","https://openalex.org/W6697234459","https://openalex.org/W6714171909","https://openalex.org/W6729959887","https://openalex.org/W6746451879","https://openalex.org/W6757132592","https://openalex.org/W6768798221","https://openalex.org/W6780998715"],"related_works":["https://openalex.org/W2509658457","https://openalex.org/W1856119614","https://openalex.org/W2167653113","https://openalex.org/W2291424197","https://openalex.org/W2354384964","https://openalex.org/W3002801003","https://openalex.org/W2765157699","https://openalex.org/W1815165135","https://openalex.org/W2098529290","https://openalex.org/W2097316030","https://openalex.org/W2106563808","https://openalex.org/W2029733119","https://openalex.org/W14254504","https://openalex.org/W2963727906","https://openalex.org/W3181776258","https://openalex.org/W1885745269","https://openalex.org/W2184012128","https://openalex.org/W2902572630","https://openalex.org/W2940347029","https://openalex.org/W2044684117"],"abstract_inverted_index":{"Deep":[0],"Neural":[1],"Network\u2013Hidden":[2],"Markov":[3],"Model":[4],"(DNN-HMM)":[5],"based":[6],"methods":[7],"have":[8],"been":[9],"successfully":[10],"used":[11],"for":[12,101,124,147,154],"many":[13],"always-on":[14],"keyword":[15,52,103,149],"spotting":[16,104],"algorithms":[17],"that":[18,117],"detect":[19],"a":[20,25,34,112],"wake":[21],"word":[22],"to":[23,49,67,143],"trigger":[24,198],"device.":[26],"The":[27,55,80],"DNN":[28,43,84,120,204],"predicts":[29],"the":[30,42,51,64,69,73,76,83,89,95,102,119,125,133,145,148,152,166,170,195],"state":[31,78],"probabilities":[32],"of":[33,45,63,98],"given":[35],"speech":[36,47,156],"frame,":[37],"while":[38],"HMM":[39,65,134],"decoder":[40,135],"combines":[41],"predictions":[44],"multiple":[46],"frames":[48],"compute":[50,182],"detection":[53,126],"score.":[54,127],"DNN,":[56],"in":[57,165,178,189],"prior":[58],"methods,":[59],"is":[60,94,175],"trained":[61],"independent":[62,203],"parameters":[66,121],"minimize":[68,151],"cross-entropy":[70],"loss":[71,86],"between":[72,82],"predicted":[74],"and":[75,88,139,150],"ground-truth":[77],"probabilities.":[79],"mismatch":[81,110],"training":[85,115],"(cross-entropy)":[87],"end":[90],"metric":[91],"(detection":[92],"score)":[93],"main":[96],"source":[97],"sub-optimal":[99],"performance":[100],"task.":[105],"We":[106],"address":[107],"this":[108,129],"loss-metric":[109],"with":[111],"novel":[113],"end-to-end":[114],"strategy":[116],"learns":[118],"by":[122],"optimizing":[123],"To":[128],"end,":[130],"we":[131,185],"make":[132],"(dynamic":[136],"programming)":[137],"differentiable":[138],"back-propagate":[140],"through":[141],"it":[142],"maximize":[144],"score":[146],"scores":[153],"non-keyword":[155],"segments.":[157],"Our":[158],"method":[159],"does":[160],"not":[161],"require":[162],"any":[163],"change":[164],"model":[167],"architecture":[168],"or":[169,181],"inference":[171],"framework;":[172],"there-fore,":[173],"there":[174],"no":[176],"overhead":[177],"run-time":[179],"memory":[180],"requirements.":[183],"Moreover,":[184],"show":[186],"significant":[187],"reduction":[188],"false":[190,197],"rejection":[191],"rate":[192],"(FRR)":[193],"at":[194],"same":[196],"experience":[199],"(>":[200],"70%":[201],"over":[202],"training).":[205]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}