{"id":"https://openalex.org/W4406461904","doi":"https://doi.org/10.1109/slt61566.2024.10832227","title":"Effective Text Adaptation For LLM-Based ASR Through Soft Prompt Fine-Tuning","display_name":"Effective Text Adaptation For LLM-Based ASR Through Soft Prompt Fine-Tuning","publication_year":2024,"publication_date":"2024-12-02","ids":{"openalex":"https://openalex.org/W4406461904","doi":"https://doi.org/10.1109/slt61566.2024.10832227"},"language":"en","primary_location":{"id":"doi:10.1109/slt61566.2024.10832227","is_oa":false,"landing_page_url":"https://doi.org/10.1109/slt61566.2024.10832227","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":"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/A5103207430","display_name":"Yingyi Ma","orcid":"https://orcid.org/0000-0002-0205-0197"},"institutions":[{"id":"https://openalex.org/I3197470489","display_name":"Alpha Omega Alpha Medical Honor Society","ror":"https://ror.org/057q9nn35","country_code":"US","type":"nonprofit","lineage":["https://openalex.org/I3197470489"]},{"id":"https://openalex.org/I4210099336","display_name":"Menlo School","ror":"https://ror.org/01240pn49","country_code":"US","type":"education","lineage":["https://openalex.org/I4210099336"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Yingyi Ma","raw_affiliation_strings":["Meta AI,Menlo Park,CA,USA"],"affiliations":[{"raw_affiliation_string":"Meta AI,Menlo Park,CA,USA","institution_ids":["https://openalex.org/I3197470489","https://openalex.org/I4210099336"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100462347","display_name":"Zhe Liu","orcid":"https://orcid.org/0000-0001-8578-2635"},"institutions":[{"id":"https://openalex.org/I4210099336","display_name":"Menlo School","ror":"https://ror.org/01240pn49","country_code":"US","type":"education","lineage":["https://openalex.org/I4210099336"]},{"id":"https://openalex.org/I3197470489","display_name":"Alpha Omega Alpha Medical Honor Society","ror":"https://ror.org/057q9nn35","country_code":"US","type":"nonprofit","lineage":["https://openalex.org/I3197470489"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhe Liu","raw_affiliation_strings":["Meta AI,Menlo Park,CA,USA"],"affiliations":[{"raw_affiliation_string":"Meta AI,Menlo Park,CA,USA","institution_ids":["https://openalex.org/I3197470489","https://openalex.org/I4210099336"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5066166549","display_name":"Ozlem Kalinli","orcid":null},"institutions":[{"id":"https://openalex.org/I3197470489","display_name":"Alpha Omega Alpha Medical Honor Society","ror":"https://ror.org/057q9nn35","country_code":"US","type":"nonprofit","lineage":["https://openalex.org/I3197470489"]},{"id":"https://openalex.org/I4210099336","display_name":"Menlo School","ror":"https://ror.org/01240pn49","country_code":"US","type":"education","lineage":["https://openalex.org/I4210099336"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ozlem Kalinli","raw_affiliation_strings":["Meta AI,Menlo Park,CA,USA"],"affiliations":[{"raw_affiliation_string":"Meta AI,Menlo Park,CA,USA","institution_ids":["https://openalex.org/I3197470489","https://openalex.org/I4210099336"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5103207430"],"corresponding_institution_ids":["https://openalex.org/I3197470489","https://openalex.org/I4210099336"],"apc_list":null,"apc_paid":null,"fwci":1.4513,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":{"value":0.85886751,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":97,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"64","last_page":"69"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9977999925613403,"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":0.9977999925613403,"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":0.9846000075340271,"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/T12326","display_name":"Network Packet Processing and Optimization","score":0.9807999730110168,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/adaptation","display_name":"Adaptation (eye)","score":0.7096149325370789},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6902320981025696},{"id":"https://openalex.org/keywords/fine-tuning","display_name":"Fine-tuning","score":0.5408779382705688},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.3391430974006653},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.32593047618865967},{"id":"https://openalex.org/keywords/psychology","display_name":"Psychology","score":0.16219162940979004},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.09915918111801147},{"id":"https://openalex.org/keywords/neuroscience","display_name":"Neuroscience","score":0.09803003072738647}],"concepts":[{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.7096149325370789},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6902320981025696},{"id":"https://openalex.org/C157524613","wikidata":"https://www.wikidata.org/wiki/Q2828883","display_name":"Fine-tuning","level":2,"score":0.5408779382705688},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.3391430974006653},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.32593047618865967},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.16219162940979004},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.09915918111801147},{"id":"https://openalex.org/C169760540","wikidata":"https://www.wikidata.org/wiki/Q207011","display_name":"Neuroscience","level":1,"score":0.09803003072738647},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/slt61566.2024.10832227","is_oa":false,"landing_page_url":"https://doi.org/10.1109/slt61566.2024.10832227","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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":40,"referenced_works":["https://openalex.org/W2107223151","https://openalex.org/W2938947737","https://openalex.org/W2963739817","https://openalex.org/W3097777922","https://openalex.org/W3163605596","https://openalex.org/W3174770825","https://openalex.org/W3196642073","https://openalex.org/W3196869722","https://openalex.org/W3197661863","https://openalex.org/W3202419788","https://openalex.org/W3205788551","https://openalex.org/W3207222250","https://openalex.org/W3208049241","https://openalex.org/W4205991051","https://openalex.org/W4223622550","https://openalex.org/W4224930731","https://openalex.org/W4225272718","https://openalex.org/W4225319488","https://openalex.org/W4226120743","https://openalex.org/W4308028428","https://openalex.org/W4319862418","https://openalex.org/W4319862474","https://openalex.org/W4322718191","https://openalex.org/W4380994269","https://openalex.org/W4385823377","https://openalex.org/W4386555501","https://openalex.org/W4386655647","https://openalex.org/W4391021666","https://openalex.org/W4391709581","https://openalex.org/W4392903956","https://openalex.org/W6800480908","https://openalex.org/W6803092890","https://openalex.org/W6845995170","https://openalex.org/W6850625674","https://openalex.org/W6852800892","https://openalex.org/W6853094705","https://openalex.org/W6854388276","https://openalex.org/W6855594702","https://openalex.org/W6856794988","https://openalex.org/W6860809563"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"The":[0],"advent":[1],"of":[2,35,79],"Large":[3],"Language":[4,142],"Models":[5],"(LLM)":[6],"has":[7],"reformed":[8],"the":[9,23,63,77,130,135,149],"Automatic":[10],"Speech":[11],"Recognition":[12],"(ASR).":[13],"Prompting":[14],"LLM":[15,64],"with":[16,31,105,140],"audio":[17],"embeddings":[18],"to":[19,113,122,134],"generate":[20],"transcriptions":[21],"becomes":[22],"new":[24],"state-of-the-art":[25],"ASR.":[26,137],"Despite":[27],"LLMs":[28],"being":[29],"trained":[30],"an":[32],"extensive":[33],"amount":[34],"text":[36,40,59,97,103],"corpora,":[37],"high-quality":[38],"domain-specific":[39,80,96,141],"data":[41,71],"can":[42,55,146],"still":[43],"significantly":[44],"enhance":[45],"ASR":[46,54,68],"performance":[47],"on":[48,69,129],"domain":[49,132],"adaptation":[50,104],"tasks.":[51],"Although":[52],"LLM-based":[53],"naturally":[56],"incorporate":[57],"more":[58],"corpora":[60],"by":[61,151],"fine-tuning":[62,66,92],"decoder,":[65],"such":[67],"text-only":[70],"without":[72],"paired":[73],"prompts":[74],"may":[75],"diminish":[76],"effectiveness":[78],"knowledge.":[81],"To":[82],"mitigate":[83],"this":[84,139],"issue,":[85],"we":[86],"propose":[87],"a":[88,110,152],"two-step":[89],"soft":[90],"prompt":[91],"strategy":[93],"that":[94,102],"enhances":[95],"adaptation.":[98],"Experimental":[99],"results":[100],"show":[101],"our":[106],"proposed":[107],"method":[108],"achieved":[109],"relative":[111,153],"up":[112,121],"9%":[114],"Word":[115],"Error":[116,125],"Rate":[117,126],"(WER)":[118],"reduction":[119,128],"and":[120],"18%":[123],"Entity":[124],"(EER)":[127],"target":[131],"compared":[133],"baseline":[136],"Combining":[138],"Model":[143],"(LM)":[144],"fusion":[145],"further":[147],"improve":[148],"EER":[150],"2-5%.":[154]},"counts_by_year":[{"year":2025,"cited_by_count":4}],"updated_date":"2026-02-25T23:00:34.991745","created_date":"2025-10-10T00:00:00"}