{"id":"https://openalex.org/W4372342913","doi":"https://doi.org/10.1109/icassp49357.2023.10095589","title":"Improving Transformer-Based End-to-End Speaker Diarization by Assigning Auxiliary Losses to Attention Heads","display_name":"Improving Transformer-Based End-to-End Speaker Diarization by Assigning Auxiliary Losses to Attention Heads","publication_year":2023,"publication_date":"2023-05-05","ids":{"openalex":"https://openalex.org/W4372342913","doi":"https://doi.org/10.1109/icassp49357.2023.10095589"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49357.2023.10095589","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49357.2023.10095589","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2023 - 2023 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/A5090642537","display_name":"Ye-Rin Jeoung","orcid":null},"institutions":[{"id":"https://openalex.org/I4575257","display_name":"Hanyang University","ror":"https://ror.org/046865y68","country_code":"KR","type":"education","lineage":["https://openalex.org/I4575257"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Ye-Rin Jeoung","raw_affiliation_strings":["Hanyang University,Department of Electronic Engineering,Seoul,Republic of Korea","Department of Electronic Engineering, Hanyang University, Seoul, Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hanyang University,Department of Electronic Engineering,Seoul,Republic of Korea","institution_ids":["https://openalex.org/I4575257"]},{"raw_affiliation_string":"Department of Electronic Engineering, Hanyang University, Seoul, Republic of Korea","institution_ids":["https://openalex.org/I4575257"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5110972432","display_name":"Joon\u2010Young Yang","orcid":null},"institutions":[{"id":"https://openalex.org/I4575257","display_name":"Hanyang University","ror":"https://ror.org/046865y68","country_code":"KR","type":"education","lineage":["https://openalex.org/I4575257"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Joon-Young Yang","raw_affiliation_strings":["Hanyang University,Department of Electronic Engineering,Seoul,Republic of Korea","Department of Electronic Engineering, Hanyang University, Seoul, Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hanyang University,Department of Electronic Engineering,Seoul,Republic of Korea","institution_ids":["https://openalex.org/I4575257"]},{"raw_affiliation_string":"Department of Electronic Engineering, Hanyang University, Seoul, Republic of Korea","institution_ids":["https://openalex.org/I4575257"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101610186","display_name":"Jeong-Hwan Choi","orcid":"https://orcid.org/0000-0002-3937-2333"},"institutions":[{"id":"https://openalex.org/I4575257","display_name":"Hanyang University","ror":"https://ror.org/046865y68","country_code":"KR","type":"education","lineage":["https://openalex.org/I4575257"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jeong-Hwan Choi","raw_affiliation_strings":["Hanyang University,Department of Electronic Engineering,Seoul,Republic of Korea","Department of Electronic Engineering, Hanyang University, Seoul, Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hanyang University,Department of Electronic Engineering,Seoul,Republic of Korea","institution_ids":["https://openalex.org/I4575257"]},{"raw_affiliation_string":"Department of Electronic Engineering, Hanyang University, Seoul, Republic of Korea","institution_ids":["https://openalex.org/I4575257"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5002418613","display_name":"Joon\u2010Hyuk Chang","orcid":"https://orcid.org/0000-0003-2610-2323"},"institutions":[{"id":"https://openalex.org/I4575257","display_name":"Hanyang University","ror":"https://ror.org/046865y68","country_code":"KR","type":"education","lineage":["https://openalex.org/I4575257"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Joon-Hyuk Chang","raw_affiliation_strings":["Hanyang University,Department of Electronic Engineering,Seoul,Republic of Korea","Department of Electronic Engineering, Hanyang University, Seoul, Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hanyang University,Department of Electronic Engineering,Seoul,Republic of Korea","institution_ids":["https://openalex.org/I4575257"]},{"raw_affiliation_string":"Department of Electronic Engineering, Hanyang University, Seoul, Republic of Korea","institution_ids":["https://openalex.org/I4575257"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I4575257"],"apc_list":null,"apc_paid":null,"fwci":1.2904,"has_fulltext":false,"cited_by_count":8,"citation_normalized_percentile":{"value":0.83674852,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":97,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"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":0.9991999864578247,"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/T11309","display_name":"Music 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"}}],"keywords":[{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.7722364664077759},{"id":"https://openalex.org/keywords/speaker-diarisation","display_name":"Speaker diarisation","score":0.7509781718254089},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7263014316558838},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.6395705342292786},{"id":"https://openalex.org/keywords/end-to-end-principle","display_name":"End-to-end principle","score":0.5357538461685181},{"id":"https://openalex.org/keywords/speaker-recognition","display_name":"Speaker recognition","score":0.5233230590820312},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.3707316219806671},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3624304533004761},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.11000606417655945}],"concepts":[{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.7722364664077759},{"id":"https://openalex.org/C149838564","wikidata":"https://www.wikidata.org/wiki/Q7574248","display_name":"Speaker diarisation","level":3,"score":0.7509781718254089},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7263014316558838},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.6395705342292786},{"id":"https://openalex.org/C74296488","wikidata":"https://www.wikidata.org/wiki/Q2527392","display_name":"End-to-end principle","level":2,"score":0.5357538461685181},{"id":"https://openalex.org/C133892786","wikidata":"https://www.wikidata.org/wiki/Q1145189","display_name":"Speaker recognition","level":2,"score":0.5233230590820312},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3707316219806671},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3624304533004761},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.11000606417655945},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.0},{"id":"https://openalex.org/C119599485","wikidata":"https://www.wikidata.org/wiki/Q43035","display_name":"Electrical engineering","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49357.2023.10095589","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49357.2023.10095589","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4","score":0.5}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":29,"referenced_works":["https://openalex.org/W119864726","https://openalex.org/W1522301498","https://openalex.org/W1591607137","https://openalex.org/W1876378865","https://openalex.org/W2060729502","https://openalex.org/W2115599677","https://openalex.org/W2159094433","https://openalex.org/W2166637769","https://openalex.org/W2170579896","https://openalex.org/W2187693774","https://openalex.org/W2219249508","https://openalex.org/W2395619215","https://openalex.org/W2792643794","https://openalex.org/W2972712416","https://openalex.org/W3008357631","https://openalex.org/W3015783745","https://openalex.org/W3144086690","https://openalex.org/W3157070662","https://openalex.org/W3198482860","https://openalex.org/W3206479921","https://openalex.org/W4297841469","https://openalex.org/W4385245566","https://openalex.org/W6631190155","https://openalex.org/W6639431793","https://openalex.org/W6688816777","https://openalex.org/W6739901393","https://openalex.org/W6749029207","https://openalex.org/W6770434673","https://openalex.org/W6780226713"],"related_works":["https://openalex.org/W2162158162","https://openalex.org/W1521049138","https://openalex.org/W1509309911","https://openalex.org/W2144208207","https://openalex.org/W2175373321","https://openalex.org/W4319663602","https://openalex.org/W4323073035","https://openalex.org/W4372342913","https://openalex.org/W3025798172","https://openalex.org/W3015923889"],"abstract_inverted_index":{"Transformer-based":[0],"end-to-end":[1],"neural":[2],"speaker":[3,16,83],"diarization":[4,142,146],"(EEND)":[5],"models":[6],"utilize":[7],"the":[8,28,36,42,46,53,71,99,105,114,119,123,129,135,145,150],"multi-head":[9],"self-attention":[10],"(SA)":[11],"mechanism":[12],"to":[13,26,68,80,87,103,108],"enable":[14],"accurate":[15],"label":[17],"prediction":[18],"in":[19,113,122],"overlapped":[20,92],"speech":[21,93],"regions.":[22],"In":[23],"this":[24],"study,":[25],"enhance":[27],"training":[29],"effectiveness":[30,127],"of":[31,38,45,57,70,128,149],"SA-EEND":[32,152],"models,":[33],"we":[34,50,97],"propose":[35],"use":[37],"auxiliary":[39,101],"losses":[40,102],"for":[41,140],"SA":[43,59,124],"heads":[44],"transformer":[47,106],"layers.":[48],"Specifically,":[49],"assume":[51],"that":[52],"attention":[54,115],"weight":[55],"matrices":[56,79],"an":[58],"layer":[60],"are":[61,66],"redundant":[62],"if":[63],"their":[64],"patterns":[65,85,112],"similar":[67],"those":[69],"identity":[72],"matrix.":[73],"We":[74],"then":[75],"explicitly":[76],"constrain":[77],"such":[78],"exhibit":[81,109],"specific":[82],"activity":[84,89],"relevant":[86],"voice":[88],"detection":[90,94],"or":[91],"tasks.":[95],"Consequently,":[96],"expect":[98],"proposed":[100,130],"guide":[104],"layers":[107],"more":[110],"diverse":[111],"weights,":[116],"thereby":[117],"reducing":[118,144],"assumed":[120],"redundancies":[121],"heads.":[125],"The":[126],"method":[131],"is":[132],"demonstrated":[133],"using":[134],"simulated":[136],"and":[137,156],"CALLHOME":[138],"datasets":[139],"two-speaker":[141],"tasks,":[143],"error":[147],"rate":[148],"conventional":[151],"model":[153],"by":[154],"32.58%":[155],"17.11%,":[157],"respectively.":[158]},"counts_by_year":[{"year":2025,"cited_by_count":4},{"year":2024,"cited_by_count":4}],"updated_date":"2026-06-26T08:34:08.712188","created_date":"2025-10-10T00:00:00"}