{"id":"https://openalex.org/W3163169798","doi":"https://doi.org/10.1109/icassp39728.2021.9414920","title":"Confidence Estimation for Attention-Based Sequence-to-Sequence Models for Speech Recognition","display_name":"Confidence Estimation for Attention-Based Sequence-to-Sequence Models for Speech Recognition","publication_year":2021,"publication_date":"2021-05-13","ids":{"openalex":"https://openalex.org/W3163169798","doi":"https://doi.org/10.1109/icassp39728.2021.9414920","mag":"3163169798"},"language":"en","primary_location":{"id":"doi:10.1109/icassp39728.2021.9414920","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp39728.2021.9414920","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":"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/A5030757951","display_name":"Qiujia Li","orcid":"https://orcid.org/0000-0003-3074-3692"},"institutions":[{"id":"https://openalex.org/I241749","display_name":"University of Cambridge","ror":"https://ror.org/013meh722","country_code":"GB","type":"education","lineage":["https://openalex.org/I241749"]}],"countries":["GB"],"is_corresponding":true,"raw_author_name":"Qiujia Li","raw_affiliation_strings":["University of Cambridge,UK","University of Cambridge, UK"],"affiliations":[{"raw_affiliation_string":"University of Cambridge,UK","institution_ids":["https://openalex.org/I241749"]},{"raw_affiliation_string":"University of Cambridge, UK","institution_ids":["https://openalex.org/I241749"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5079902855","display_name":"David Qiu","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":"David Qiu","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100433648","display_name":"Yu Zhang","orcid":"https://orcid.org/0000-0002-9505-1833"},"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":"Yu Zhang","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100677404","display_name":"Bo Li","orcid":"https://orcid.org/0000-0003-1483-8077"},"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":"Bo Li","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101319167","display_name":"Yanzhang He","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":"Yanzhang He","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5002191410","display_name":"Philip C. Woodland","orcid":"https://orcid.org/0000-0001-9069-0225"},"institutions":[{"id":"https://openalex.org/I241749","display_name":"University of Cambridge","ror":"https://ror.org/013meh722","country_code":"GB","type":"education","lineage":["https://openalex.org/I241749"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Philip C. Woodland","raw_affiliation_strings":["University of Cambridge,UK","University of Cambridge, UK"],"affiliations":[{"raw_affiliation_string":"University of Cambridge,UK","institution_ids":["https://openalex.org/I241749"]},{"raw_affiliation_string":"University of Cambridge, UK","institution_ids":["https://openalex.org/I241749"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103187717","display_name":"Liangliang Cao","orcid":"https://orcid.org/0000-0003-0900-1512"},"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":"Liangliang Cao","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5032931723","display_name":"Trevor Strohman","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":"Trevor Strohman","raw_affiliation_strings":["Google LLC,USA","Google LLC, USA"],"affiliations":[{"raw_affiliation_string":"Google LLC,USA","institution_ids":["https://openalex.org/I1291425158"]},{"raw_affiliation_string":"Google LLC, USA","institution_ids":["https://openalex.org/I1291425158"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":8,"corresponding_author_ids":["https://openalex.org/A5030757951"],"corresponding_institution_ids":["https://openalex.org/I241749"],"apc_list":null,"apc_paid":null,"fwci":5.21,"has_fulltext":false,"cited_by_count":44,"citation_normalized_percentile":{"value":0.96165448,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":97,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"6388","last_page":"6392"},"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9990000128746033,"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":0.9975000023841858,"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/softmax-function","display_name":"Softmax function","score":0.8146527409553528},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7466352581977844},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.7070149183273315},{"id":"https://openalex.org/keywords/hidden-markov-model","display_name":"Hidden Markov model","score":0.6376094222068787},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.6225740313529968},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.6112366914749146},{"id":"https://openalex.org/keywords/sequence","display_name":"Sequence (biology)","score":0.5580713152885437},{"id":"https://openalex.org/keywords/overconfidence-effect","display_name":"Overconfidence effect","score":0.49898624420166016},{"id":"https://openalex.org/keywords/sequence-labeling","display_name":"Sequence labeling","score":0.4651912748813629},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4431169927120209},{"id":"https://openalex.org/keywords/word","display_name":"Word (group theory)","score":0.42531150579452515},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.1958751380443573},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.16505539417266846},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.15104025602340698},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.09926354885101318}],"concepts":[{"id":"https://openalex.org/C188441871","wikidata":"https://www.wikidata.org/wiki/Q7554146","display_name":"Softmax function","level":3,"score":0.8146527409553528},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7466352581977844},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.7070149183273315},{"id":"https://openalex.org/C23224414","wikidata":"https://www.wikidata.org/wiki/Q176769","display_name":"Hidden Markov model","level":2,"score":0.6376094222068787},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.6225740313529968},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.6112366914749146},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.5580713152885437},{"id":"https://openalex.org/C51110983","wikidata":"https://www.wikidata.org/wiki/Q16503490","display_name":"Overconfidence effect","level":2,"score":0.49898624420166016},{"id":"https://openalex.org/C35639132","wikidata":"https://www.wikidata.org/wiki/Q7452468","display_name":"Sequence labeling","level":3,"score":0.4651912748813629},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4431169927120209},{"id":"https://openalex.org/C90805587","wikidata":"https://www.wikidata.org/wiki/Q10944557","display_name":"Word (group theory)","level":2,"score":0.42531150579452515},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.1958751380443573},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.16505539417266846},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.15104025602340698},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.09926354885101318},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","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},{"id":"https://openalex.org/C54355233","wikidata":"https://www.wikidata.org/wiki/Q7162","display_name":"Genetics","level":1,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.0},{"id":"https://openalex.org/C77805123","wikidata":"https://www.wikidata.org/wiki/Q161272","display_name":"Social psychology","level":1,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp39728.2021.9414920","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp39728.2021.9414920","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"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Peace, Justice and strong institutions","id":"https://metadata.un.org/sdg/16","score":0.5899999737739563}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":52,"referenced_works":["https://openalex.org/W34303869","https://openalex.org/W173561343","https://openalex.org/W854541894","https://openalex.org/W1494198834","https://openalex.org/W1795658042","https://openalex.org/W1828163288","https://openalex.org/W1904457459","https://openalex.org/W1915251500","https://openalex.org/W1976526581","https://openalex.org/W2024490156","https://openalex.org/W2034537249","https://openalex.org/W2046566693","https://openalex.org/W2086161653","https://openalex.org/W2095629250","https://openalex.org/W2095705004","https://openalex.org/W2108677974","https://openalex.org/W2121879602","https://openalex.org/W2129334286","https://openalex.org/W2130799511","https://openalex.org/W2134659216","https://openalex.org/W2143577772","https://openalex.org/W2159391028","https://openalex.org/W2161391345","https://openalex.org/W2183341477","https://openalex.org/W2280141299","https://openalex.org/W2327501763","https://openalex.org/W2531327146","https://openalex.org/W2594610113","https://openalex.org/W2626967530","https://openalex.org/W2898630520","https://openalex.org/W2928941594","https://openalex.org/W2936774411","https://openalex.org/W2950170869","https://openalex.org/W2962760690","https://openalex.org/W2962824709","https://openalex.org/W2963505832","https://openalex.org/W2964212410","https://openalex.org/W2972953886","https://openalex.org/W3007905516","https://openalex.org/W3015870704","https://openalex.org/W3026041220","https://openalex.org/W4234016251","https://openalex.org/W6607114211","https://openalex.org/W6623517193","https://openalex.org/W6638749077","https://openalex.org/W6640059789","https://openalex.org/W6674330103","https://openalex.org/W6676315081","https://openalex.org/W6679825567","https://openalex.org/W6728622933","https://openalex.org/W6739651123","https://openalex.org/W6760633627"],"related_works":["https://openalex.org/W2991634017","https://openalex.org/W3107204728","https://openalex.org/W4287591324","https://openalex.org/W2535098683","https://openalex.org/W2268150819","https://openalex.org/W2437596814","https://openalex.org/W2085373452","https://openalex.org/W4287601428","https://openalex.org/W4205820553","https://openalex.org/W2251441308"],"abstract_inverted_index":{"For":[0],"various":[1],"speech-related":[2],"tasks,":[3],"confidence":[4,30,91,110,138],"scores":[5,31,92,139],"from":[6,36,158],"a":[7,11,104,146,159],"speech":[8,26,157],"recogniser":[9],"are":[10],"useful":[12],"measure":[13],"to":[14,66,156],"assess":[15],"the":[16,68,73,89,95,130],"quality":[17],"of":[18,98,116,145],"transcriptions.":[19],"In":[20,76],"traditional":[21],"hidden":[22],"Markov":[23],"model-based":[24],"automatic":[25],"recognition":[27],"(ASR)":[28],"systems,":[29],"can":[32,128,134,164],"be":[33],"reliably":[34],"obtained":[35],"word":[37,58],"posteriors":[38,59],"in":[39],"decoding":[40],"lattices.":[41],"However,":[42],"for":[43],"an":[44,48,53,117],"ASR":[45,120],"system":[46],"with":[47,140],"auto-regressive":[49],"decoder,":[50],"such":[51,169],"as":[52,72,170],"attention-based":[54],"sequence-to-sequence":[55],"model,":[56],"computing":[57],"is":[60,65],"difficult.":[61],"An":[62],"obvious":[63],"alternative":[64],"use":[67],"decoder":[69],"softmax":[70],"probability":[71],"model":[74],"confidence.":[75],"this":[77],"paper,":[78],"we":[79,102],"first":[80],"examine":[81],"how":[82],"some":[83],"commonly":[84],"used":[85],"regularisation":[86],"methods":[87],"influence":[88],"softmax-based":[90],"and":[93,106,133,141,163],"study":[94],"overconfident":[96],"behaviour":[97],"end-to-end":[99,119],"models.":[100],"Then":[101],"propose":[103],"lightweight":[105],"effective":[107],"approach":[108],"named":[109],"estimation":[111],"module":[112],"(CEM)":[113],"on":[114,123],"top":[115],"existing":[118],"model.":[121,148],"Experiments":[122],"LibriSpeech":[124],"show":[125],"that":[126,152],"CEM":[127,153],"mitigate":[129],"overconfidence":[131],"problem":[132],"produce":[135],"more":[136],"reliable":[137],"without":[142],"shallow":[143],"fusion":[144],"language":[147],"Further":[149],"analysis":[150],"shows":[151],"generalises":[154],"well":[155],"moderately":[160],"mismatched":[161],"domain":[162],"potentially":[165],"improve":[166],"downstream":[167],"tasks":[168],"semi-supervised":[171],"learning.":[172]},"counts_by_year":[{"year":2025,"cited_by_count":7},{"year":2024,"cited_by_count":8},{"year":2023,"cited_by_count":13},{"year":2022,"cited_by_count":5},{"year":2021,"cited_by_count":11}],"updated_date":"2026-04-03T22:45:19.894376","created_date":"2025-10-10T00:00:00"}