{"id":"https://openalex.org/W4296068780","doi":"https://doi.org/10.21437/interspeech.2022-239","title":"Tokenwise Contrastive Pretraining for Finer Speech-to-BERT Alignment in End-to-End Speech-to-Intent Systems","display_name":"Tokenwise Contrastive Pretraining for Finer Speech-to-BERT Alignment in End-to-End Speech-to-Intent Systems","publication_year":2022,"publication_date":"2022-09-16","ids":{"openalex":"https://openalex.org/W4296068780","doi":"https://doi.org/10.21437/interspeech.2022-239"},"language":"en","primary_location":{"id":"doi:10.21437/interspeech.2022-239","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2022-239","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":"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/A5113145191","display_name":"Vishal Sunder","orcid":null},"institutions":[{"id":"https://openalex.org/I52357470","display_name":"The Ohio State University","ror":"https://ror.org/00rs6vg23","country_code":"US","type":"education","lineage":["https://openalex.org/I52357470"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Vishal Sunder","raw_affiliation_strings":["The Ohio State University, Columbus, OH, USA"],"affiliations":[{"raw_affiliation_string":"The Ohio State University, Columbus, OH, USA","institution_ids":["https://openalex.org/I52357470"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5056667180","display_name":"Eric Fosler\u2010Lussier","orcid":"https://orcid.org/0000-0001-8004-5169"},"institutions":[{"id":"https://openalex.org/I52357470","display_name":"The Ohio State University","ror":"https://ror.org/00rs6vg23","country_code":"US","type":"education","lineage":["https://openalex.org/I52357470"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Eric Fosler-Lussier","raw_affiliation_strings":["The Ohio State University, Columbus, OH, USA"],"affiliations":[{"raw_affiliation_string":"The Ohio State University, Columbus, OH, USA","institution_ids":["https://openalex.org/I52357470"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101787514","display_name":"Samuel Thomas","orcid":"https://orcid.org/0000-0001-7573-0620"},"institutions":[{"id":"https://openalex.org/I1341412227","display_name":"IBM (United States)","ror":"https://ror.org/05hh8d621","country_code":"US","type":"company","lineage":["https://openalex.org/I1341412227"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Samuel Thomas","raw_affiliation_strings":["IBM Research AI, Yorktown Heights, NY, USA"],"affiliations":[{"raw_affiliation_string":"IBM Research AI, Yorktown Heights, NY, USA","institution_ids":["https://openalex.org/I1341412227"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5110252428","display_name":"Hong-Kwang Jeff Kuo","orcid":null},"institutions":[{"id":"https://openalex.org/I1341412227","display_name":"IBM (United States)","ror":"https://ror.org/05hh8d621","country_code":"US","type":"company","lineage":["https://openalex.org/I1341412227"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Hong-Kwang Kuo","raw_affiliation_strings":["IBM Research AI, Yorktown Heights, NY, USA"],"affiliations":[{"raw_affiliation_string":"IBM Research AI, Yorktown Heights, NY, USA","institution_ids":["https://openalex.org/I1341412227"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5003725957","display_name":"Brian Kingsbury","orcid":"https://orcid.org/0000-0002-1343-6837"},"institutions":[{"id":"https://openalex.org/I1341412227","display_name":"IBM (United States)","ror":"https://ror.org/05hh8d621","country_code":"US","type":"company","lineage":["https://openalex.org/I1341412227"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Brian Kingsbury","raw_affiliation_strings":["IBM Research AI, Yorktown Heights, NY, USA"],"affiliations":[{"raw_affiliation_string":"IBM Research AI, Yorktown Heights, NY, USA","institution_ids":["https://openalex.org/I1341412227"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5113145191"],"corresponding_institution_ids":["https://openalex.org/I52357470"],"apc_list":null,"apc_paid":null,"fwci":0.5218,"has_fulltext":false,"cited_by_count":5,"citation_normalized_percentile":{"value":0.63698993,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"2683","last_page":"2687"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12031","display_name":"Speech and dialogue systems","score":0.998199999332428,"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/T12031","display_name":"Speech and dialogue systems","score":0.998199999332428,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.9754999876022339,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.960099995136261,"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/end-to-end-principle","display_name":"End-to-end principle","score":0.8681939840316772},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7260704636573792},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.6197463870048523},{"id":"https://openalex.org/keywords/speech-translation","display_name":"Speech translation","score":0.5183587670326233},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.28564080595970154},{"id":"https://openalex.org/keywords/machine-translation","display_name":"Machine translation","score":0.13286206126213074}],"concepts":[{"id":"https://openalex.org/C74296488","wikidata":"https://www.wikidata.org/wiki/Q2527392","display_name":"End-to-end principle","level":2,"score":0.8681939840316772},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7260704636573792},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.6197463870048523},{"id":"https://openalex.org/C2780366754","wikidata":"https://www.wikidata.org/wiki/Q7494857","display_name":"Speech translation","level":3,"score":0.5183587670326233},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.28564080595970154},{"id":"https://openalex.org/C203005215","wikidata":"https://www.wikidata.org/wiki/Q79798","display_name":"Machine translation","level":2,"score":0.13286206126213074}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.21437/interspeech.2022-239","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2022-239","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":[{"id":"https://openalex.org/G3193728207","display_name":"RI: Small: Early Elementary Reading Verification in Challenging Acoustic Environments","funder_award_id":"2008043","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G848032724","display_name":null,"funder_award_id":"Science","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":27,"referenced_works":["https://openalex.org/W1494198834","https://openalex.org/W2327501763","https://openalex.org/W2896457183","https://openalex.org/W2914417638","https://openalex.org/W2936774411","https://openalex.org/W2938630833","https://openalex.org/W2965373594","https://openalex.org/W2972584841","https://openalex.org/W3005680577","https://openalex.org/W3016006013","https://openalex.org/W3036601975","https://openalex.org/W3096813338","https://openalex.org/W3147670539","https://openalex.org/W3148001440","https://openalex.org/W3161302809","https://openalex.org/W3162313915","https://openalex.org/W3162646066","https://openalex.org/W3169320628","https://openalex.org/W3174321708","https://openalex.org/W3176896803","https://openalex.org/W3202591724","https://openalex.org/W4200113972","https://openalex.org/W4226031872","https://openalex.org/W4287238261","https://openalex.org/W4287596060","https://openalex.org/W4292779060","https://openalex.org/W4385245566"],"related_works":["https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W3179968364","https://openalex.org/W2990025607","https://openalex.org/W3045103338","https://openalex.org/W3007142233","https://openalex.org/W3037186962","https://openalex.org/W3177132412","https://openalex.org/W2989819126","https://openalex.org/W4372260133"],"abstract_inverted_index":{"Recent":[0],"advances":[1],"in":[2,47],"End-to-End":[3],"(E2E)":[4],"Spoken":[5],"Language":[6],"Understanding":[7],"(SLU)":[8],"have":[9],"been":[10],"primarily":[11],"due":[12],"to":[13,34,79,115],"effective":[14],"pretraining":[15,20],"of":[16,25],"speech":[17,35,58,86,120,143],"representations.One":[18],"such":[19,88,111],"paradigm":[21],"is":[22,40,103,144],"the":[23,45],"distillation":[24],"semantic":[26],"knowledge":[27],"from":[28,84],"state-of-the-art":[29,123],"text-based":[30],"models":[31],"like":[32],"BERT":[33,61,98],"encoder":[36,87],"neural":[37],"networks.This":[38],"work":[39],"a":[41,48,64,68,75,85,106,112],"step":[42],"towards":[43],"doing":[44],"same":[46],"much":[49],"more":[50],"efficient":[51],"and":[52,60,95],"fine-grained":[53],"manner":[54],"where":[55],"we":[56],"align":[57],"embeddings":[59,62,83],"on":[63,125],"token-by-token":[65],"basis.We":[66],"introduce":[67],"simple":[69],"yet":[70],"novel":[71,107],"technique":[72],"that":[73,89],"uses":[74],"cross-modal":[76],"attention":[77],"mechanism":[78],"extract":[80],"token-level":[81],"contextual":[82,100],"these":[90],"can":[91],"be":[92],"directly":[93,121],"compared":[94],"aligned":[96],"with":[97,136],"based":[99],"embeddings.This":[101],"alignment":[102],"performed":[104],"using":[105,119,139],"tokenwise":[108],"contrastive":[109],"loss.Fine-tuning":[110],"pretrained":[113],"model":[114,131],"perform":[116],"intent":[117],"recognition":[118],"yields":[122],"performance":[124],"two":[126],"widely":[127],"used":[128],"SLU":[129],"datasets.Our":[130],"improves":[132],"further":[133],"when":[134,142],"fine-tuned":[135],"additional":[137],"regularization":[138],"SpecAugment":[140],"especially":[141],"noisy,":[145],"giving":[146],"an":[147],"absolute":[148],"improvement":[149],"as":[150,152],"high":[151],"8%":[153],"over":[154],"previous":[155],"results.":[156]},"counts_by_year":[{"year":2024,"cited_by_count":3},{"year":2023,"cited_by_count":2}],"updated_date":"2026-04-10T15:06:20.359241","created_date":"2025-10-10T00:00:00"}