{"id":"https://openalex.org/W3176579201","doi":"https://doi.org/10.1109/taslp.2021.3092567","title":"A Novel Loss Function and Training Strategy for Noise-Robust Keyword Spotting","display_name":"A Novel Loss Function and Training Strategy for Noise-Robust Keyword Spotting","publication_year":2021,"publication_date":"2021-01-01","ids":{"openalex":"https://openalex.org/W3176579201","doi":"https://doi.org/10.1109/taslp.2021.3092567","mag":"3176579201"},"language":"en","primary_location":{"id":"doi:10.1109/taslp.2021.3092567","is_oa":false,"landing_page_url":"https://doi.org/10.1109/taslp.2021.3092567","pdf_url":null,"source":{"id":"https://openalex.org/S4210169297","display_name":"IEEE/ACM Transactions on Audio Speech and Language Processing","issn_l":"2329-9290","issn":["2329-9290","2329-9304"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE/ACM Transactions on Audio, Speech, and Language Processing","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://vbn.aau.dk/ws/files/532712423/A_Novel_Loss_Function_and_Training_Strategy_for_Noise_Robust_Keyword_Spotting.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5048592470","display_name":"Iv\u00e1n L\u00f3pez\u2010Espejo","orcid":"https://orcid.org/0000-0001-8634-7897"},"institutions":[{"id":"https://openalex.org/I891191580","display_name":"Aalborg University","ror":"https://ror.org/04m5j1k67","country_code":"DK","type":"education","lineage":["https://openalex.org/I891191580"]}],"countries":["DK"],"is_corresponding":false,"raw_author_name":"Iv\u00e1n L\u00f3pez-Espejo","raw_affiliation_strings":["Department of Electronic Systems, Aalborg University, Aalborg, Denmark"],"raw_orcid":"https://orcid.org/0000-0001-8634-7897","affiliations":[{"raw_affiliation_string":"Department of Electronic Systems, Aalborg University, Aalborg, Denmark","institution_ids":["https://openalex.org/I891191580"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5090108098","display_name":"Zheng\u2010Hua Tan","orcid":"https://orcid.org/0000-0001-6856-8928"},"institutions":[{"id":"https://openalex.org/I891191580","display_name":"Aalborg University","ror":"https://ror.org/04m5j1k67","country_code":"DK","type":"education","lineage":["https://openalex.org/I891191580"]}],"countries":["DK"],"is_corresponding":false,"raw_author_name":"Zheng-Hua Tan","raw_affiliation_strings":["Department of Electronic Systems, Aalborg University, Aalborg, Denmark"],"raw_orcid":"https://orcid.org/0000-0001-6856-8928","affiliations":[{"raw_affiliation_string":"Department of Electronic Systems, Aalborg University, Aalborg, Denmark","institution_ids":["https://openalex.org/I891191580"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5101679595","display_name":"Jesper Jensen","orcid":"https://orcid.org/0000-0003-1478-622X"},"institutions":[{"id":"https://openalex.org/I4210130492","display_name":"Oticon Medical (Denmark)","ror":"https://ror.org/03c5ds320","country_code":"DK","type":"company","lineage":["https://openalex.org/I4210130492"]},{"id":"https://openalex.org/I891191580","display_name":"Aalborg University","ror":"https://ror.org/04m5j1k67","country_code":"DK","type":"education","lineage":["https://openalex.org/I891191580"]}],"countries":["DK"],"is_corresponding":false,"raw_author_name":"Jesper Jensen","raw_affiliation_strings":["Department of Electronic Systems, Aalborg University, Aalborg, Denmark","Oticon A/S, Sm\u00f8rum, Denmark"],"raw_orcid":"https://orcid.org/0000-0003-1478-622X","affiliations":[{"raw_affiliation_string":"Department of Electronic Systems, Aalborg University, Aalborg, Denmark","institution_ids":["https://openalex.org/I891191580"]},{"raw_affiliation_string":"Oticon A/S, Sm\u00f8rum, Denmark","institution_ids":["https://openalex.org/I4210130492"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":3.0785,"has_fulltext":true,"cited_by_count":27,"citation_normalized_percentile":{"value":0.92801559,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":95,"max":99},"biblio":{"volume":"29","issue":null,"first_page":"2254","last_page":"2266"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T13083","display_name":"Advanced Text Analysis 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"}},"topics":[{"id":"https://openalex.org/T13083","display_name":"Advanced Text Analysis 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/T11800","display_name":"User Authentication and Security Systems","score":0.9937000274658203,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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":0.9873999953269958,"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/keyword-spotting","display_name":"Keyword spotting","score":0.8371810913085938},{"id":"https://openalex.org/keywords/spotting","display_name":"Spotting","score":0.7087770700454712},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.6003542542457581},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.5485086441040039},{"id":"https://openalex.org/keywords/function","display_name":"Function (biology)","score":0.5143064260482788},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.490818589925766},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.37228256464004517},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3527076244354248},{"id":"https://openalex.org/keywords/geography","display_name":"Geography","score":0.13903328776359558},{"id":"https://openalex.org/keywords/biology","display_name":"Biology","score":0.11869409680366516}],"concepts":[{"id":"https://openalex.org/C2781213101","wikidata":"https://www.wikidata.org/wiki/Q6398558","display_name":"Keyword spotting","level":2,"score":0.8371810913085938},{"id":"https://openalex.org/C2779506182","wikidata":"https://www.wikidata.org/wiki/Q7580141","display_name":"Spotting","level":2,"score":0.7087770700454712},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.6003542542457581},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5485086441040039},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.5143064260482788},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.490818589925766},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.37228256464004517},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3527076244354248},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.13903328776359558},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.11869409680366516},{"id":"https://openalex.org/C153294291","wikidata":"https://www.wikidata.org/wiki/Q25261","display_name":"Meteorology","level":1,"score":0.0},{"id":"https://openalex.org/C78458016","wikidata":"https://www.wikidata.org/wiki/Q840400","display_name":"Evolutionary biology","level":1,"score":0.0},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.0}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.1109/taslp.2021.3092567","is_oa":false,"landing_page_url":"https://doi.org/10.1109/taslp.2021.3092567","pdf_url":null,"source":{"id":"https://openalex.org/S4210169297","display_name":"IEEE/ACM Transactions on Audio Speech and Language Processing","issn_l":"2329-9290","issn":["2329-9290","2329-9304"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE/ACM Transactions on Audio, Speech, and Language Processing","raw_type":"journal-article"},{"id":"pmh:oai:pure.atira.dk:publications/bf7d0c41-12e1-4c8e-b5f0-90b2d1198b8e","is_oa":true,"landing_page_url":"https://vbn.aau.dk/da/publications/bf7d0c41-12e1-4c8e-b5f0-90b2d1198b8e","pdf_url":"https://vbn.aau.dk/ws/files/532712423/A_Novel_Loss_Function_and_Training_Strategy_for_Noise_Robust_Keyword_Spotting.pdf","source":{"id":"https://openalex.org/S4306401731","display_name":"VBN Forskningsportal (Aalborg Universitet)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I891191580","host_organization_name":"Aalborg University","host_organization_lineage":["https://openalex.org/I891191580"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Espejo, I L, Tan, Z-H & Jensen, J 2021, 'A Novel Loss Function and Training Strategy for Noise-Robust Keyword Spotting', IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 29, 9465680, pp. 2254 - 2266. https://doi.org/10.1109/TASLP.2021.3092567","raw_type":"info:eu-repo/semantics/publishedVersion"},{"id":"pmh:oai:pure.atira.dk:publications/bf7d0c41-12e1-4c8e-b5f0-90b2d1198b8e","is_oa":true,"landing_page_url":"http://www.scopus.com/inward/record.url?scp=85110759492&partnerID=8YFLogxK","pdf_url":null,"source":{"id":"https://openalex.org/S4306401731","display_name":"VBN Forskningsportal (Aalborg Universitet)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I891191580","host_organization_name":"Aalborg University","host_organization_lineage":["https://openalex.org/I891191580"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Espejo , I L , Tan , Z-H &amp; Jensen , J 2021 , ' A Novel Loss Function and Training Strategy for Noise-Robust Keyword Spotting ' , IEEE/ACM Transactions on Audio, Speech, and Language Processing , vol. 29 , 9465680 , pp. 2254 - 2266 . https://doi.org/10.1109/TASLP.2021.3092567","raw_type":"article"}],"best_oa_location":{"id":"pmh:oai:pure.atira.dk:publications/bf7d0c41-12e1-4c8e-b5f0-90b2d1198b8e","is_oa":true,"landing_page_url":"https://vbn.aau.dk/da/publications/bf7d0c41-12e1-4c8e-b5f0-90b2d1198b8e","pdf_url":"https://vbn.aau.dk/ws/files/532712423/A_Novel_Loss_Function_and_Training_Strategy_for_Noise_Robust_Keyword_Spotting.pdf","source":{"id":"https://openalex.org/S4306401731","display_name":"VBN Forskningsportal (Aalborg Universitet)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I891191580","host_organization_name":"Aalborg University","host_organization_lineage":["https://openalex.org/I891191580"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Espejo, I L, Tan, Z-H & Jensen, J 2021, 'A Novel Loss Function and Training Strategy for Noise-Robust Keyword Spotting', IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 29, 9465680, pp. 2254 - 2266. https://doi.org/10.1109/TASLP.2021.3092567","raw_type":"info:eu-repo/semantics/publishedVersion"},"sustainable_development_goals":[{"score":0.46000000834465027,"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4"}],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W3176579201.pdf","grobid_xml":"https://content.openalex.org/works/W3176579201.grobid-xml"},"referenced_works_count":68,"referenced_works":["https://openalex.org/W8870360","https://openalex.org/W183625566","https://openalex.org/W1522301498","https://openalex.org/W1553469512","https://openalex.org/W1974387177","https://openalex.org/W2096733369","https://openalex.org/W2139983529","https://openalex.org/W2141730410","https://openalex.org/W2151484683","https://openalex.org/W2157364932","https://openalex.org/W2187089797","https://openalex.org/W2289394825","https://openalex.org/W2483244395","https://openalex.org/W2555897561","https://openalex.org/W2559809918","https://openalex.org/W2598634450","https://openalex.org/W2602634800","https://openalex.org/W2606377603","https://openalex.org/W2618099328","https://openalex.org/W2767325013","https://openalex.org/W2769912137","https://openalex.org/W2783089003","https://openalex.org/W2787051072","https://openalex.org/W2797583228","https://openalex.org/W2888641632","https://openalex.org/W2889017397","https://openalex.org/W2890992173","https://openalex.org/W2910715461","https://openalex.org/W2937769560","https://openalex.org/W2953219395","https://openalex.org/W2962758570","https://openalex.org/W2963350250","https://openalex.org/W2963376890","https://openalex.org/W2963403868","https://openalex.org/W2963414149","https://openalex.org/W2963628261","https://openalex.org/W2963881567","https://openalex.org/W2964121744","https://openalex.org/W2964187693","https://openalex.org/W2970428702","https://openalex.org/W2972163809","https://openalex.org/W2972613398","https://openalex.org/W2973226577","https://openalex.org/W2980451698","https://openalex.org/W3005031961","https://openalex.org/W3007913012","https://openalex.org/W3008790380","https://openalex.org/W3012505451","https://openalex.org/W3015118013","https://openalex.org/W3016218307","https://openalex.org/W3025581723","https://openalex.org/W3031469774","https://openalex.org/W3099206234","https://openalex.org/W3114848746","https://openalex.org/W3148564648","https://openalex.org/W4385245566","https://openalex.org/W6600367688","https://openalex.org/W6631190155","https://openalex.org/W6682194319","https://openalex.org/W6730323794","https://openalex.org/W6735531217","https://openalex.org/W6739901393","https://openalex.org/W6746451879","https://openalex.org/W6750665317","https://openalex.org/W6773098122","https://openalex.org/W6777189519","https://openalex.org/W6777440055","https://openalex.org/W6843142645"],"related_works":["https://openalex.org/W2918559346","https://openalex.org/W2114097550","https://openalex.org/W4286904253","https://openalex.org/W3119978414","https://openalex.org/W2516975559","https://openalex.org/W3206647229","https://openalex.org/W4408353962","https://openalex.org/W1969408022","https://openalex.org/W2000885660","https://openalex.org/W2545741539"],"abstract_inverted_index":{"The":[0],"development":[1],"of":[2,40,48,105,146],"keyword":[3,50,64],"spotting":[4],"(KWS)":[5],"systems":[6],"that":[7,57,112],"are":[8,58],"accurate":[9],"in":[10,19],"noisy":[11,103],"conditions":[12],"remains":[13],"a":[14,24,49,54,70,102],"challenge.":[15],"Towards":[16],"this":[17,20,36],"goal,":[18],"paper":[21],"we":[22,38,67],"propose":[23,69],"novel":[25],"training":[26,31,128],"strategy":[27],"relying":[28],"on":[29,101],"multi-condition":[30,127],"for":[32],"noise-robust":[33],"KWS.":[34],"By":[35],"strategy,":[37],"think":[39],"the":[41,46,63,80,106,143,147],"state-of-the-art":[42],"KWS":[43,118,148],"models":[44],"as":[45],"composition":[47],"embedding":[51,65],"extractor":[52],"and":[53,88,95],"linear":[55],"classifier":[56],"successively":[59],"trained.":[60],"To":[61],"train":[62],"extractor,":[66],"also":[68],"new":[71],"(C":[72],"<sub":[73],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[74],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">N,2</sub>":[75],"+1)-pair":[76],"loss":[77,84],"function":[78],"extending":[79],"concept":[81],"behind":[82],"related":[83],"functions":[85],"like":[86],"triplet":[87],"N-pair":[89],"losses":[90],"to":[91,124],"reach":[92],"larger":[93],"inter-class":[94],"smaller":[96],"intra-class":[97],"variation.":[98],"Experimental":[99],"results":[100],"version":[104],"Google":[107],"Speech":[108],"Commands":[109],"Dataset":[110],"show":[111],"our":[113],"proposal":[114],"achieves":[115],"around":[116],"12%":[117],"accuracy":[119],"relative":[120],"improvement":[121,138],"with":[122],"respect":[123],"standard":[125],"end-to-end":[126],"when":[129],"speech":[130],"is":[131,139],"distorted":[132],"by":[133],"unseen":[134],"noises.":[135],"This":[136],"performance":[137],"achieved":[140],"without":[141],"increasing":[142],"computational":[144],"complexity":[145],"model.":[149]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":4},{"year":2024,"cited_by_count":7},{"year":2023,"cited_by_count":5},{"year":2022,"cited_by_count":7},{"year":2021,"cited_by_count":3}],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}