{"id":"https://openalex.org/W4412445775","doi":"https://doi.org/10.1109/noms57970.2025.11073667","title":"Resource-Efficient Federated Fine-Tuning of LLMs for Downstream Tasks of Speech Recognition","display_name":"Resource-Efficient Federated Fine-Tuning of LLMs for Downstream Tasks of Speech Recognition","publication_year":2025,"publication_date":"2025-05-12","ids":{"openalex":"https://openalex.org/W4412445775","doi":"https://doi.org/10.1109/noms57970.2025.11073667"},"language":"en","primary_location":{"id":"doi:10.1109/noms57970.2025.11073667","is_oa":false,"landing_page_url":"https://doi.org/10.1109/noms57970.2025.11073667","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"NOMS 2025-2025 IEEE Network Operations and Management Symposium","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/A5051639604","display_name":"Shivani Sanjay Kolekar","orcid":"https://orcid.org/0000-0001-8422-6071"},"institutions":[{"id":"https://openalex.org/I111277659","display_name":"Chonnam National University","ror":"https://ror.org/05kzjxq56","country_code":"KR","type":"education","lineage":["https://openalex.org/I111277659"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"Shivani Sanjay Kolekar","raw_affiliation_strings":["Chonnam National University,Department of Artificial Intelligence Convergence,Gwangju,South Korea"],"affiliations":[{"raw_affiliation_string":"Chonnam National University,Department of Artificial Intelligence Convergence,Gwangju,South Korea","institution_ids":["https://openalex.org/I111277659"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5015500546","display_name":"Kyungbaek Kim","orcid":"https://orcid.org/0000-0001-9985-3051"},"institutions":[{"id":"https://openalex.org/I111277659","display_name":"Chonnam National University","ror":"https://ror.org/05kzjxq56","country_code":"KR","type":"education","lineage":["https://openalex.org/I111277659"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Kyungbaek Kim","raw_affiliation_strings":["Chonnam National University,Department of Artificial Intelligence Convergence,Gwangju,South Korea"],"affiliations":[{"raw_affiliation_string":"Chonnam National University,Department of Artificial Intelligence Convergence,Gwangju,South Korea","institution_ids":["https://openalex.org/I111277659"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5051639604"],"corresponding_institution_ids":["https://openalex.org/I111277659"],"apc_list":null,"apc_paid":null,"fwci":2.5513,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.90712332,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":91,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.9988999962806702,"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.9988999962806702,"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/T10237","display_name":"Cryptography and Data Security","score":0.996399998664856,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9915000200271606,"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/downstream","display_name":"Downstream (manufacturing)","score":0.7127076983451843},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6515893936157227},{"id":"https://openalex.org/keywords/resource","display_name":"Resource (disambiguation)","score":0.5263023376464844},{"id":"https://openalex.org/keywords/fine-tuning","display_name":"Fine-tuning","score":0.4708901047706604},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.43971046805381775},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.17768850922584534},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.11012771725654602},{"id":"https://openalex.org/keywords/operations-management","display_name":"Operations management","score":0.08310690522193909}],"concepts":[{"id":"https://openalex.org/C2776207758","wikidata":"https://www.wikidata.org/wiki/Q5303302","display_name":"Downstream (manufacturing)","level":2,"score":0.7127076983451843},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6515893936157227},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.5263023376464844},{"id":"https://openalex.org/C157524613","wikidata":"https://www.wikidata.org/wiki/Q2828883","display_name":"Fine-tuning","level":2,"score":0.4708901047706604},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.43971046805381775},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.17768850922584534},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.11012771725654602},{"id":"https://openalex.org/C21547014","wikidata":"https://www.wikidata.org/wiki/Q1423657","display_name":"Operations management","level":1,"score":0.08310690522193909},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/noms57970.2025.11073667","is_oa":false,"landing_page_url":"https://doi.org/10.1109/noms57970.2025.11073667","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"NOMS 2025-2025 IEEE Network Operations and Management Symposium","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":29,"referenced_works":["https://openalex.org/W179875071","https://openalex.org/W2158195707","https://openalex.org/W2560674852","https://openalex.org/W2763421725","https://openalex.org/W2783652918","https://openalex.org/W2786771851","https://openalex.org/W2888781539","https://openalex.org/W2896457183","https://openalex.org/W2963318081","https://openalex.org/W2963819344","https://openalex.org/W2997958863","https://openalex.org/W3035190488","https://openalex.org/W3129831491","https://openalex.org/W3161051256","https://openalex.org/W3204960814","https://openalex.org/W4214607237","https://openalex.org/W4293649366","https://openalex.org/W4311000453","https://openalex.org/W4322717062","https://openalex.org/W4372338301","https://openalex.org/W4385245566","https://openalex.org/W4385848613","https://openalex.org/W4389708319","https://openalex.org/W4391020683","https://openalex.org/W4400925671","https://openalex.org/W4401023469","https://openalex.org/W4402754051","https://openalex.org/W4402754082","https://openalex.org/W4406461754"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W1583765404","https://openalex.org/W4214653257","https://openalex.org/W2390279801","https://openalex.org/W2055438207","https://openalex.org/W4391913857","https://openalex.org/W2521424917","https://openalex.org/W2358668433"],"abstract_inverted_index":{"Fine-tuning":[0],"large":[1],"pre-trained":[2],"speech":[3],"recognition":[4],"models":[5],"on":[6,109,121,153,167],"resource-constrained":[7,122,155],"edge":[8,168],"devices":[9],"within":[10],"federated":[11,165],"learning":[12,166],"frameworks":[13],"presents":[14],"significant":[15],"challenges":[16],"due":[17],"to":[18,40,80,142],"computational":[19,116,133],"limitations,":[20],"communication":[21],"overhead,":[22],"and":[23,60,84,104,118,132,145,161],"the":[24,49,54,57,74,86,89,95,110],"complexities":[25],"of":[26,56,76,92],"mixed-precision":[27,69,107],"training.":[28],"To":[29],"address":[30],"these":[31],"issues,":[32],"we":[33,52,98],"propose":[34],"FedMP-Head,":[35],"a":[36,63],"two-step":[37],"method":[38],"designed":[39],"optimize":[41],"training":[42,108],"efficiency":[43],"while":[44],"maintaining":[45],"high":[46],"accuracy.":[47],"In":[48,94],"first":[50],"step,":[51,97],"freeze":[53],"encoder":[55],"foundation":[58],"model":[59,78],"train":[61],"only":[62],"lightweight":[64],"classifier":[65],"head":[66],"before":[67],"applying":[68],"quantization.":[70],"This":[71,113],"approach":[72],"reduces":[73],"size":[75],"global":[77],"updates":[79],"approximately":[81],"780K":[82],"parameters":[83],"maintains":[85],"accuracy":[87,148],"during":[88],"downstream":[90],"task":[91],"classification.":[93],"second":[96],"implement":[99],"budgeted":[100],"layer-wise":[101],"gradient":[102],"scaling":[103],"correction":[105],"with":[106,129],"client":[111],"side.":[112],"technique":[114],"optimizes":[115],"cost":[117],"memory":[119],"footprint":[120],"clients":[123,156],"by":[124],"effectively":[125],"utilizing":[126],"gradients,":[127],"even":[128],"limited":[130],"precision":[131],"capabilities.":[134],"FedMP-Head":[135],"accelerates":[136],"training,":[137],"reducing":[138],"convergence":[139],"time":[140],"compared":[141],"full-precision":[143],"methods":[144],"achieving":[146],"threshold":[147],"in":[149],"fewer":[150],"rounds.":[151],"Experiments":[152],"100":[154],"highlight":[157],"its":[158],"improved":[159],"performance":[160],"suitability":[162],"for":[163],"real-world":[164],"devices.":[169]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-03-25T14:56:36.534964","created_date":"2025-10-10T00:00:00"}