{"id":"https://openalex.org/W7148471489","doi":"https://doi.org/10.1109/asru65441.2025.11434678","title":"Hybrid Decoding: Rapid Pass and Selective Detailed Correction for Sequence Models","display_name":"Hybrid Decoding: Rapid Pass and Selective Detailed Correction for Sequence Models","publication_year":2025,"publication_date":"2025-12-06","ids":{"openalex":"https://openalex.org/W7148471489","doi":"https://doi.org/10.1109/asru65441.2025.11434678"},"language":null,"primary_location":{"id":"doi:10.1109/asru65441.2025.11434678","is_oa":false,"landing_page_url":"https://doi.org/10.1109/asru65441.2025.11434678","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)","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/A5027173062","display_name":"Yunkyu Lim","orcid":null},"institutions":[{"id":"https://openalex.org/I18003638","display_name":"Mando Corporation (South Korea)","ror":"https://ror.org/051v3pe92","country_code":"KR","type":"company","lineage":["https://openalex.org/I18003638"]},{"id":"https://openalex.org/I4210086063","display_name":"Handok (South Korea)","ror":"https://ror.org/00exehz38","country_code":"KR","type":"company","lineage":["https://openalex.org/I4210086063"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"Yunkyu Lim","raw_affiliation_strings":["42dot Inc.,,Republic of Korea"],"affiliations":[{"raw_affiliation_string":"42dot Inc.,,Republic of Korea","institution_ids":["https://openalex.org/I18003638","https://openalex.org/I4210086063"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100740584","display_name":"Jihwan Park","orcid":"https://orcid.org/0009-0007-2223-7278"},"institutions":[{"id":"https://openalex.org/I18003638","display_name":"Mando Corporation (South Korea)","ror":"https://ror.org/051v3pe92","country_code":"KR","type":"company","lineage":["https://openalex.org/I18003638"]},{"id":"https://openalex.org/I4210086063","display_name":"Handok (South Korea)","ror":"https://ror.org/00exehz38","country_code":"KR","type":"company","lineage":["https://openalex.org/I4210086063"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jihwan Park","raw_affiliation_strings":["42dot Inc.,,Republic of Korea"],"affiliations":[{"raw_affiliation_string":"42dot Inc.,,Republic of Korea","institution_ids":["https://openalex.org/I18003638","https://openalex.org/I4210086063"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132808834","display_name":"Hyung Yong Kim","orcid":null},"institutions":[{"id":"https://openalex.org/I18003638","display_name":"Mando Corporation (South Korea)","ror":"https://ror.org/051v3pe92","country_code":"KR","type":"company","lineage":["https://openalex.org/I18003638"]},{"id":"https://openalex.org/I4210086063","display_name":"Handok (South Korea)","ror":"https://ror.org/00exehz38","country_code":"KR","type":"company","lineage":["https://openalex.org/I4210086063"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Hyung Yong Kim","raw_affiliation_strings":["42dot Inc.,,Republic of Korea"],"affiliations":[{"raw_affiliation_string":"42dot Inc.,,Republic of Korea","institution_ids":["https://openalex.org/I18003638","https://openalex.org/I4210086063"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5025565566","display_name":"Hanbin Lee","orcid":"https://orcid.org/0000-0002-4545-0027"},"institutions":[{"id":"https://openalex.org/I18003638","display_name":"Mando Corporation (South Korea)","ror":"https://ror.org/051v3pe92","country_code":"KR","type":"company","lineage":["https://openalex.org/I18003638"]},{"id":"https://openalex.org/I4210086063","display_name":"Handok (South Korea)","ror":"https://ror.org/00exehz38","country_code":"KR","type":"company","lineage":["https://openalex.org/I4210086063"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Hanbin Lee","raw_affiliation_strings":["42dot Inc.,,Republic of Korea"],"affiliations":[{"raw_affiliation_string":"42dot Inc.,,Republic of Korea","institution_ids":["https://openalex.org/I18003638","https://openalex.org/I4210086063"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5053784244","display_name":"Byeong-Yeol Kim","orcid":"https://orcid.org/0000-0001-6019-5047"},"institutions":[{"id":"https://openalex.org/I18003638","display_name":"Mando Corporation (South Korea)","ror":"https://ror.org/051v3pe92","country_code":"KR","type":"company","lineage":["https://openalex.org/I18003638"]},{"id":"https://openalex.org/I4210086063","display_name":"Handok (South Korea)","ror":"https://ror.org/00exehz38","country_code":"KR","type":"company","lineage":["https://openalex.org/I4210086063"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Byeong-Yeol Kim","raw_affiliation_strings":["42dot Inc.,,Republic of Korea"],"affiliations":[{"raw_affiliation_string":"42dot Inc.,,Republic of Korea","institution_ids":["https://openalex.org/I18003638","https://openalex.org/I4210086063"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5027173062"],"corresponding_institution_ids":["https://openalex.org/I18003638","https://openalex.org/I4210086063"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.87569943,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.2881999909877777,"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":0.2881999909877777,"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/T11269","display_name":"Algorithms and Data Compression","score":0.11550000309944153,"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/T12072","display_name":"Machine Learning and Algorithms","score":0.062199998646974564,"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/sequence","display_name":"Sequence (biology)","score":0.5425000190734863},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.2962999939918518},{"id":"https://openalex.org/keywords/filter","display_name":"Filter (signal processing)","score":0.26080000400543213},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.257099986076355},{"id":"https://openalex.org/keywords/stability","display_name":"Stability (learning theory)","score":0.2522999942302704}],"concepts":[{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.5425000190734863},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5076000094413757},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.44679999351501465},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3147999942302704},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.2962999939918518},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2799000144004822},{"id":"https://openalex.org/C106131492","wikidata":"https://www.wikidata.org/wiki/Q3072260","display_name":"Filter (signal processing)","level":2,"score":0.26080000400543213},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.257099986076355},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.2522999942302704},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.24860000610351562}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/asru65441.2025.11434678","is_oa":false,"landing_page_url":"https://doi.org/10.1109/asru65441.2025.11434678","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)","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":9,"referenced_works":["https://openalex.org/W1494198834","https://openalex.org/W2127141656","https://openalex.org/W2327501763","https://openalex.org/W2952030198","https://openalex.org/W2963747784","https://openalex.org/W3198694222","https://openalex.org/W4319862631","https://openalex.org/W4391021542","https://openalex.org/W4402111700"],"related_works":[],"abstract_inverted_index":{"Recently,":[0],"Transformer-based":[1],"encoder-decoder":[2,62],"models":[3],"have":[4],"demonstrated":[5],"strong":[6],"performance":[7],"in":[8,98],"multilingual":[9],"speech":[10],"recognition.":[11],"However,":[12],"the":[13,53,60,71,76,93,109,121,135,141],"decoder\u2019s":[14],"autoregressive":[15],"nature":[16],"and":[17,31,51,101,111],"large":[18],"size":[19],"introduce":[20],"significant":[21],"bottlenecks":[22],"during":[23],"inference.":[24],"Additionally,":[25],"although":[26],"rare,":[27],"repetition":[28],"can":[29],"occur":[30],"negatively":[32],"affect":[33],"recognition":[34],"accuracy.":[35],"To":[36],"tackle":[37],"these":[38],"challenges,":[39],"we":[40],"propose":[41],"a":[42,66],"novel":[43],"Hybrid":[44],"Decoding":[45],"approach":[46],"that":[47],"both":[48],"accelerates":[49],"inference":[50,142],"alleviates":[52],"issue":[54],"of":[55],"repetition.":[56],"Our":[57],"method":[58,125],"extends":[59],"transformer":[61],"architecture":[63],"by":[64,92],"attaching":[65],"lightweight,":[67],"fast":[68,77],"decoder":[69,78],"to":[70,120,131],"pretrained":[72],"encoder.":[73],"During":[74],"inference,":[75],"rapidly":[79],"generates":[80],"an":[81],"output,":[82],"which":[83],"is":[84],"then":[85],"verified":[86],"and,":[87],"if":[88],"necessary,":[89],"selectively":[90],"corrected":[91],"Transformer":[94],"decoder.":[95],"This":[96],"results":[97],"faster":[99],"decoding":[100],"improved":[102],"robustness":[103],"against":[104],"repetitive":[105],"errors.":[106],"Experiments":[107],"on":[108],"LibriSpeech":[110],"GigaSpeech":[112],"test":[113],"sets":[114],"indicate":[115],"that,":[116],"with":[117],"finetuning":[118],"limited":[119],"added":[122],"decoder,":[123],"our":[124],"achieves":[126],"word":[127],"error":[128],"rates":[129],"comparable":[130],"or":[132],"better":[133],"than":[134,139],"baseline,":[136],"while":[137],"more":[138],"doubling":[140],"speed.":[143]},"counts_by_year":[],"updated_date":"2026-04-03T16:44:17.987007","created_date":"2026-04-03T00:00:00"}