{"id":"https://openalex.org/W4408354168","doi":"https://doi.org/10.1109/icassp49660.2025.10889444","title":"Developing Instruction-Following Speech Language Model Without Speech Instruction-Tuning Data","display_name":"Developing Instruction-Following Speech Language Model Without Speech Instruction-Tuning Data","publication_year":2025,"publication_date":"2025-03-12","ids":{"openalex":"https://openalex.org/W4408354168","doi":"https://doi.org/10.1109/icassp49660.2025.10889444"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49660.2025.10889444","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10889444","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","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/A5015216145","display_name":"Ke-Han Lu","orcid":null},"institutions":[{"id":"https://openalex.org/I16733864","display_name":"National Taiwan University","ror":"https://ror.org/05bqach95","country_code":"TW","type":"education","lineage":["https://openalex.org/I16733864"]}],"countries":["TW"],"is_corresponding":false,"raw_author_name":"Ke-Han Lu","raw_affiliation_strings":["National Taiwan University,Graduate Institute of Communication Engineering"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"National Taiwan University,Graduate Institute of Communication Engineering","institution_ids":["https://openalex.org/I16733864"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5002433660","display_name":"Zhehuai Chen","orcid":"https://orcid.org/0000-0003-4400-5340"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhehuai Chen","raw_affiliation_strings":["NVIDIA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071471469","display_name":"Szu\u2010Wei Fu","orcid":"https://orcid.org/0000-0002-3487-8212"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Szu-Wei Fu","raw_affiliation_strings":["NVIDIA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5020376803","display_name":"Chao-Han Huck Yang","orcid":"https://orcid.org/0000-0003-2879-8811"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chao-Han Huck Yang","raw_affiliation_strings":["NVIDIA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5040747392","display_name":"Jagadeesh Balam","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jagadeesh Balam","raw_affiliation_strings":["NVIDIA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5032957280","display_name":"Boris Ginsburg","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Boris Ginsburg","raw_affiliation_strings":["NVIDIA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5090045508","display_name":"Yu-Chiang Frank Wang","orcid":"https://orcid.org/0000-0002-2333-157X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu-Chiang Frank Wang","raw_affiliation_strings":["NVIDIA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5040508737","display_name":"Hung-yi Lee","orcid":"https://orcid.org/0000-0002-9654-5747"},"institutions":[{"id":"https://openalex.org/I16733864","display_name":"National Taiwan University","ror":"https://ror.org/05bqach95","country_code":"TW","type":"education","lineage":["https://openalex.org/I16733864"]}],"countries":["TW"],"is_corresponding":false,"raw_author_name":"Hung-Yi Lee","raw_affiliation_strings":["National Taiwan University,Graduate Institute of Communication Engineering"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"National Taiwan University,Graduate Institute of Communication Engineering","institution_ids":["https://openalex.org/I16733864"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":13,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"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/T12031","display_name":"Speech and dialogue systems","score":0.9072999954223633,"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/T12031","display_name":"Speech and dialogue systems","score":0.9072999954223633,"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/computer-science","display_name":"Computer science","score":0.8284560441970825},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5241149663925171},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.47958460450172424},{"id":"https://openalex.org/keywords/speech-synthesis","display_name":"Speech synthesis","score":0.47525259852409363},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.37761643528938293}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8284560441970825},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5241149663925171},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.47958460450172424},{"id":"https://openalex.org/C14999030","wikidata":"https://www.wikidata.org/wiki/Q16346","display_name":"Speech synthesis","level":2,"score":0.47525259852409363},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.37761643528938293}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49660.2025.10889444","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10889444","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","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":33,"referenced_works":["https://openalex.org/W2146334809","https://openalex.org/W2726515241","https://openalex.org/W2979826702","https://openalex.org/W4372260474","https://openalex.org/W4375869257","https://openalex.org/W4387865047","https://openalex.org/W4391021627","https://openalex.org/W4392902623","https://openalex.org/W4393178509","https://openalex.org/W4401042284","https://openalex.org/W4402112110","https://openalex.org/W4402669711","https://openalex.org/W4402684170","https://openalex.org/W4403653439","https://openalex.org/W4404781855","https://openalex.org/W4404782655","https://openalex.org/W4406461677","https://openalex.org/W6767671539","https://openalex.org/W6772349387","https://openalex.org/W6777335856","https://openalex.org/W6796581206","https://openalex.org/W6809646742","https://openalex.org/W6847363464","https://openalex.org/W6849177959","https://openalex.org/W6852418670","https://openalex.org/W6852800892","https://openalex.org/W6853249747","https://openalex.org/W6856227084","https://openalex.org/W6857054612","https://openalex.org/W6859099255","https://openalex.org/W6866575372","https://openalex.org/W6870775139","https://openalex.org/W6871553294"],"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":{"Recent":[0],"end-to-end":[1],"speech":[2,19,27,34,71,99,158],"language":[3,13,50,81],"models":[4,14],"(SLMs)":[5],"have":[6],"expanded":[7],"upon":[8],"the":[9,31,48,79,84,96,114,136,151],"capabilities":[10,82,91],"of":[11,47,83,140],"large":[12],"(LLMs)":[15],"by":[16],"incorporating":[17],"pre-trained":[18],"models.":[20],"However,":[21],"these":[22],"SLMs":[23,76,141],"often":[24],"undergo":[25],"extensive":[26,147],"instruction-tuning":[28,100],"to":[29,116],"bridge":[30],"gap":[32],"between":[33],"and":[35,43,107,128,138,156],"text":[36],"modalities.":[37],"This":[38],"requires":[39],"significant":[40],"annotation":[41],"efforts":[42],"risks":[44],"catastrophic":[45],"forgetting":[46],"original":[49],"capabilities.":[51],"In":[52],"this":[53],"work,":[54],"we":[55],"present":[56],"a":[57],"simple":[58],"yet":[59],"effective":[60],"automatic":[61],"process":[62],"for":[63,92,98,153],"creating":[64],"speech-text":[65],"pair":[66],"data":[67],"that":[68],"carefully":[69],"injects":[70],"paralinguistic":[72],"understanding":[73,159],"abilities":[74],"into":[75],"while":[77],"preserving":[78],"inherent":[80],"text-based":[85],"LLM.":[86],"Our":[87,131],"model":[88,112],"demonstrates":[89],"general":[90],"speech-related":[93],"tasks":[94],"without":[95],"need":[97],"data,":[101],"achieving":[102],"impressive":[103],"performance":[104],"on":[105,146],"Dynamic-SUPERB":[106],"AIR-Bench-Chat":[108],"benchmarks.":[109],"Furthermore,":[110],"our":[111],"exhibits":[113],"ability":[115],"follow":[117],"complex":[118],"instructions":[119],"derived":[120],"from":[121],"LLMs,":[122],"such":[123],"as":[124],"specific":[125],"output":[126],"formatting":[127],"chain-of-thought":[129],"reasoning.":[130],"approach":[132],"not":[133],"only":[134],"enhances":[135],"versatility":[137],"effectiveness":[139],"but":[142],"also":[143],"reduces":[144],"reliance":[145],"annotated":[148],"datasets,":[149],"paving":[150],"way":[152],"more":[154],"efficient":[155],"capable":[157],"systems.":[160],"<sup":[161],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[162],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">1</sup>":[163]},"counts_by_year":[{"year":2026,"cited_by_count":5},{"year":2025,"cited_by_count":8}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-03-13T00:00:00"}