{"id":"https://openalex.org/W4406461721","doi":"https://doi.org/10.1109/slt61566.2024.10832209","title":"Token-Weighted RNN-T For Learning From Flawed Data","display_name":"Token-Weighted RNN-T For Learning From Flawed Data","publication_year":2024,"publication_date":"2024-12-02","ids":{"openalex":"https://openalex.org/W4406461721","doi":"https://doi.org/10.1109/slt61566.2024.10832209"},"language":"en","primary_location":{"id":"doi:10.1109/slt61566.2024.10832209","is_oa":false,"landing_page_url":"https://doi.org/10.1109/slt61566.2024.10832209","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE Spoken Language Technology Workshop (SLT)","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/A5048538280","display_name":"Gil Keren","orcid":"https://orcid.org/0000-0002-5153-3494"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gil Keren","raw_affiliation_strings":["Meta AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Meta AI","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114804421","display_name":"Wei Zhou","orcid":"https://orcid.org/0000-0001-8328-3736"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wei Zhou","raw_affiliation_strings":["Meta AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Meta AI","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5066166549","display_name":"Ozlem Kalinli","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ozlem Kalinli","raw_affiliation_strings":["Meta AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Meta AI","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.3556,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.61429729,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"22","last_page":"29"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9945999979972839,"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":0.9945999979972839,"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/T11512","display_name":"Anomaly Detection Techniques and Applications","score":0.9884999990463257,"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/T12357","display_name":"Digital Media Forensic Detection","score":0.9876999855041504,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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":0.707898736000061},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.6720240116119385},{"id":"https://openalex.org/keywords/recurrent-neural-network","display_name":"Recurrent neural network","score":0.6021554470062256},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.43208563327789307},{"id":"https://openalex.org/keywords/computer-security","display_name":"Computer security","score":0.16080543398857117},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.15889456868171692}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.707898736000061},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.6720240116119385},{"id":"https://openalex.org/C147168706","wikidata":"https://www.wikidata.org/wiki/Q1457734","display_name":"Recurrent neural network","level":3,"score":0.6021554470062256},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.43208563327789307},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.16080543398857117},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.15889456868171692}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/slt61566.2024.10832209","is_oa":false,"landing_page_url":"https://doi.org/10.1109/slt61566.2024.10832209","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE Spoken Language Technology Workshop (SLT)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":40,"referenced_works":["https://openalex.org/W569478347","https://openalex.org/W1494198834","https://openalex.org/W1588359339","https://openalex.org/W2056786202","https://openalex.org/W2085598899","https://openalex.org/W2124558353","https://openalex.org/W2160815625","https://openalex.org/W2327501763","https://openalex.org/W2608702473","https://openalex.org/W2748795451","https://openalex.org/W2962760690","https://openalex.org/W2962824709","https://openalex.org/W2963250244","https://openalex.org/W2991213871","https://openalex.org/W3008174054","https://openalex.org/W3015522062","https://openalex.org/W3095350795","https://openalex.org/W3096338464","https://openalex.org/W3096888553","https://openalex.org/W3097777922","https://openalex.org/W3162665866","https://openalex.org/W3197223534","https://openalex.org/W3197991202","https://openalex.org/W3198098585","https://openalex.org/W3204696009","https://openalex.org/W3209976096","https://openalex.org/W3213673948","https://openalex.org/W4210663600","https://openalex.org/W4220743925","https://openalex.org/W4297727296","https://openalex.org/W4372260363","https://openalex.org/W4385822648","https://openalex.org/W4386025763","https://openalex.org/W4391021555","https://openalex.org/W4391021574","https://openalex.org/W4398796251","https://openalex.org/W6687566353","https://openalex.org/W6770506093","https://openalex.org/W6809993426","https://openalex.org/W6868951638"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W4388335561","https://openalex.org/W2970530566","https://openalex.org/W4288261899","https://openalex.org/W4225394202","https://openalex.org/W4307309205","https://openalex.org/W2967478618","https://openalex.org/W4385009901"],"abstract_inverted_index":{"ASR":[0],"models":[1],"are":[2],"commonly":[3],"trained":[4],"with":[5,56,96],"the":[6,11,20,26,53,72,111,121,139],"cross-entropy":[7],"criterion":[8,50],"to":[9,34,99,105],"increase":[10],"probability":[12,21],"of":[13,22,118,138],"a":[14,46,100],"target":[15,27],"token":[16],"sequence.":[17],"While":[18],"optimizing":[19],"all":[23],"tokens":[24,36],"in":[25,71,78,120],"sequence":[28],"is":[29,62,129],"sensible,":[30],"one":[31],"may":[32],"want":[33],"de-emphasize":[35],"that":[37,51,89,126],"reflect":[38],"transcription":[39],"errors.":[40,85],"In":[41],"this":[42,133],"work,":[43],"we":[44],"propose":[45],"novel":[47],"token-weighted":[48,127],"RNN-T":[49,54,128],"augments":[52],"objective":[55,61],"token-specific":[57],"weights.":[58],"The":[59],"new":[60],"used":[63],"for":[64,93,131],"mitigating":[65],"accuracy":[66,102,112,140],"loss":[67],"from":[68,115],"transcriptions":[69],"errors":[70],"training":[73],"data,":[74],"which":[75],"naturally":[76],"appear":[77],"two":[79],"settings:":[80],"pseudo-labeling":[81],"and":[82,124],"human":[83],"annotation":[84],"Experiments":[86],"results":[87],"show":[88,125],"using":[90],"our":[91],"method":[92],"semi-supervised":[94],"learning":[95],"pseudo-labels":[97],"leads":[98],"consistent":[101],"improvement,":[103],"up":[104],"38%":[106],"relative.":[107],"We":[108],"also":[109],"analyze":[110],"degradation":[113],"resulting":[114],"different":[116],"levels":[117],"WER":[119],"reference":[122],"transcription,":[123],"suitable":[130],"overcoming":[132],"degradation,":[134],"recovering":[135],"64%":[136],"99%":[137],"loss.":[141]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}