{"id":"https://openalex.org/W4297841458","doi":"https://doi.org/10.21437/interspeech.2022-10560","title":"Analysis of Self-Attention Head Diversity for Conformer-based Automatic Speech Recognition","display_name":"Analysis of Self-Attention Head Diversity for Conformer-based Automatic Speech Recognition","publication_year":2022,"publication_date":"2022-09-16","ids":{"openalex":"https://openalex.org/W4297841458","doi":"https://doi.org/10.21437/interspeech.2022-10560"},"language":"en","primary_location":{"id":"doi:10.21437/interspeech.2022-10560","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2022-10560","pdf_url":null,"source":{"id":"https://openalex.org/S4363604309","display_name":"Interspeech 2022","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":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Interspeech 2022","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/A5015927589","display_name":"Kartik Audhkhasi","orcid":"https://orcid.org/0000-0002-2340-1144"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Kartik Audhkhasi","raw_affiliation_strings":["Google LLC, New York"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google LLC, New York","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101673889","display_name":"Yinghui Huang","orcid":"https://orcid.org/0000-0003-0607-1507"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yinghui Huang","raw_affiliation_strings":["Google LLC, New York"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google LLC, New York","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071715737","display_name":"Bhuvana Ramabhadran","orcid":null},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Bhuvana Ramabhadran","raw_affiliation_strings":["Google LLC, New York"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google LLC, New York","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5036214942","display_name":"Pedro Moreno","orcid":"https://orcid.org/0000-0002-5501-5272"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Pedro J. Moreno","raw_affiliation_strings":["Google LLC, New York"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google LLC, New York","institution_ids":["https://openalex.org/I1291425158"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I1291425158"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1026","last_page":"1030"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9994000196456909,"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":0.9994000196456909,"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":0.993399977684021,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.98089998960495,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6368280649185181},{"id":"https://openalex.org/keywords/head","display_name":"Head (geology)","score":0.6303187608718872},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5883030295372009},{"id":"https://openalex.org/keywords/diversity","display_name":"Diversity (politics)","score":0.5539000034332275},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.4317716062068939},{"id":"https://openalex.org/keywords/conformational-isomerism","display_name":"Conformational isomerism","score":0.43154770135879517},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3730234205722809},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.06505748629570007},{"id":"https://openalex.org/keywords/geology","display_name":"Geology","score":0.06358852982521057}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6368280649185181},{"id":"https://openalex.org/C2780312720","wikidata":"https://www.wikidata.org/wiki/Q5689100","display_name":"Head (geology)","level":2,"score":0.6303187608718872},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5883030295372009},{"id":"https://openalex.org/C2781316041","wikidata":"https://www.wikidata.org/wiki/Q1230584","display_name":"Diversity (politics)","level":2,"score":0.5539000034332275},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4317716062068939},{"id":"https://openalex.org/C18705241","wikidata":"https://www.wikidata.org/wiki/Q1128023","display_name":"Conformational isomerism","level":3,"score":0.43154770135879517},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3730234205722809},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.06505748629570007},{"id":"https://openalex.org/C127313418","wikidata":"https://www.wikidata.org/wiki/Q1069","display_name":"Geology","level":0,"score":0.06358852982521057},{"id":"https://openalex.org/C32909587","wikidata":"https://www.wikidata.org/wiki/Q11369","display_name":"Molecule","level":2,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C144024400","wikidata":"https://www.wikidata.org/wiki/Q21201","display_name":"Sociology","level":0,"score":0.0},{"id":"https://openalex.org/C19165224","wikidata":"https://www.wikidata.org/wiki/Q23404","display_name":"Anthropology","level":1,"score":0.0},{"id":"https://openalex.org/C114793014","wikidata":"https://www.wikidata.org/wiki/Q52109","display_name":"Geomorphology","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.21437/interspeech.2022-10560","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2022-10560","pdf_url":null,"source":{"id":"https://openalex.org/S4363604309","display_name":"Interspeech 2022","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":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Interspeech 2022","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":26,"referenced_works":["https://openalex.org/W854541894","https://openalex.org/W1494198834","https://openalex.org/W1514535095","https://openalex.org/W2133564696","https://openalex.org/W2157331557","https://openalex.org/W2327501763","https://openalex.org/W2597655663","https://openalex.org/W2604799547","https://openalex.org/W2742947407","https://openalex.org/W2804078698","https://openalex.org/W2928941594","https://openalex.org/W2946794439","https://openalex.org/W2951527505","https://openalex.org/W2962741254","https://openalex.org/W2962822108","https://openalex.org/W2963495494","https://openalex.org/W2964089206","https://openalex.org/W2970120757","https://openalex.org/W2970777192","https://openalex.org/W3016010032","https://openalex.org/W3097777922","https://openalex.org/W3106483960","https://openalex.org/W4226370720","https://openalex.org/W4288088457","https://openalex.org/W4309793872","https://openalex.org/W4385245566"],"related_works":["https://openalex.org/W2023036309","https://openalex.org/W1972863456","https://openalex.org/W2089371831","https://openalex.org/W2052399476","https://openalex.org/W2810163160","https://openalex.org/W3083970380","https://openalex.org/W2037964119","https://openalex.org/W4220905048","https://openalex.org/W4385077663","https://openalex.org/W838464343"],"abstract_inverted_index":{"Attention":[0],"layers":[1],"are":[2],"an":[3,30],"integral":[4],"part":[5,16],"of":[6,17,33,45,51,81,127,143,149,152],"modern":[7],"end-to-end":[8],"automatic":[9],"speech":[10],"recognition":[11],"systems,":[12],"for":[13,97],"instance":[14],"as":[15],"the":[18,39,52,55,60,66,73,79,133,141,147,150],"Transformer":[19],"or":[20],"Conformer":[21],"architecture.Attention":[22],"is":[23,48,118],"typically":[24],"multi-headed,":[25],"where":[26],"each":[27,98],"head":[28,90,99,107,153],"has":[29],"independent":[31],"set":[32],"learned":[34],"parameters":[35],"and":[36,70,100,123,146],"operates":[37],"on":[38,132],"same":[40],"input":[41],"feature":[42],"sequence.The":[43],"output":[44],"multi-headed":[46],"attention":[47,68,89,95,144],"a":[49,84,119,138],"fusion":[50],"outputs":[53],"from":[54],"individual":[56],"heads.We":[57],"empirically":[58],"analyze":[59],"diversity":[61,142],"between":[62,140],"representations":[63],"produced":[64],"by":[65],"different":[67,94],"heads":[69,74,145],"demonstrate":[71],"that":[72,110],"become":[75],"highly":[76],"correlated":[77],"during":[78,116],"course":[80],"training.We":[82],"investigate":[83],"few":[85],"approaches":[86],"to":[87,105,129],"increasing":[88],"diversity,":[91],"including":[92],"using":[93],"mechanisms":[96],"auxiliary":[101,113],"training":[102,117],"loss":[103,114],"functions":[104,115],"promote":[106],"diversity.We":[108],"show":[109],"introducing":[111],"diversity-promoting":[112],"more":[120],"effective":[121],"approach,":[122],"obtain":[124],"WER":[125],"improvements":[126],"up":[128],"6%":[130],"relative":[131],"Librispeech":[134],"corpus.Finally,":[135],"we":[136],"draw":[137],"connection":[139],"similarity":[148],"gradients":[151],"parameters.":[154]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":1},{"year":2023,"cited_by_count":2}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}
