{"id":"https://openalex.org/W4415433546","doi":"https://doi.org/10.21437/interspeech.2025-764","title":"Speech LLMs in Low-Resource Scenarios: Data Volume Requirements and the Impact of Pretraining on High-Resource Languages","display_name":"Speech LLMs in Low-Resource Scenarios: Data Volume Requirements and the Impact of Pretraining on High-Resource Languages","publication_year":2025,"publication_date":"2025-08-17","ids":{"openalex":"https://openalex.org/W4415433546","doi":"https://doi.org/10.21437/interspeech.2025-764"},"language":"en","primary_location":{"id":"doi:10.21437/interspeech.2025-764","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2025-764","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Interspeech 2025","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://www.isca-archive.org/interspeech_2025/fong25_interspeech.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Seraphina Fong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Seraphina Fong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037923499","display_name":"Marco Matassoni","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Marco Matassoni","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5066363315","display_name":"Alessio Brutti","orcid":"https://orcid.org/0000-0003-4146-3071"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Alessio Brutti","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":2.806,"has_fulltext":true,"cited_by_count":2,"citation_normalized_percentile":{"value":0.91801119,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":97,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"2003","last_page":"2007"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9986000061035156,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9986000061035156,"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/volume","display_name":"Volume (thermodynamics)","score":0.5170000195503235},{"id":"https://openalex.org/keywords/real-world-data","display_name":"Real world data","score":0.2207999974489212}],"concepts":[{"id":"https://openalex.org/C20556612","wikidata":"https://www.wikidata.org/wiki/Q4469374","display_name":"Volume (thermodynamics)","level":2,"score":0.5170000195503235},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.45730000734329224},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.37779998779296875},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.34700000286102295},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.34389999508857727},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.3165000081062317},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3109999895095825},{"id":"https://openalex.org/C548259974","wikidata":"https://www.wikidata.org/wiki/Q569965","display_name":"Audiology","level":1,"score":0.24549999833106995},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.22349999845027924},{"id":"https://openalex.org/C3020493868","wikidata":"https://www.wikidata.org/wiki/Q55631277","display_name":"Real world data","level":2,"score":0.2207999974489212}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.21437/interspeech.2025-764","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2025-764","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Interspeech 2025","raw_type":"proceedings-article"},{"id":"pmh:oai:iris.unitn.it:11572/465317","is_oa":true,"landing_page_url":"https://hdl.handle.net/11572/465317","pdf_url":"https://www.isca-archive.org/interspeech_2025/fong25_interspeech.pdf","source":{"id":"https://openalex.org/S4306401913","display_name":"Institutional Research Information System (Universit\u00e0 degli Studi di Trento)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I193223587","host_organization_name":"University of Trento","host_organization_lineage":["https://openalex.org/I193223587"],"host_organization_lineage_names":[],"type":"repository"},"license":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/conferenceObject"}],"best_oa_location":{"id":"pmh:oai:iris.unitn.it:11572/465317","is_oa":true,"landing_page_url":"https://hdl.handle.net/11572/465317","pdf_url":"https://www.isca-archive.org/interspeech_2025/fong25_interspeech.pdf","source":{"id":"https://openalex.org/S4306401913","display_name":"Institutional Research Information System (Universit\u00e0 degli Studi di Trento)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I193223587","host_organization_name":"University of Trento","host_organization_lineage":["https://openalex.org/I193223587"],"host_organization_lineage_names":[],"type":"repository"},"license":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/conferenceObject"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4415433546.pdf","grobid_xml":"https://content.openalex.org/works/W4415433546.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"have":[4],"demonstrated":[5],"potential":[6],"in":[7,17,27],"handling":[8],"spoken":[9],"inputs":[10],"for":[11,38,116,123],"high-resource":[12,87],"languages,":[13],"reaching":[14],"state-of-the-art":[15],"performance":[16,109],"various":[18],"tasks.":[19],"However,":[20],"their":[21],"applicability":[22],"is":[23],"still":[24],"less":[25],"explored":[26],"low-resource":[28,39,124],"settings.":[29],"This":[30],"work":[31],"investigates":[32],"the":[33,44,71,90],"use":[34],"of":[35,73,92],"Speech":[36,41,121],"LLMs":[37,102,122],"Automatic":[40],"Recognition":[42],"using":[43],"SLAM-ASR":[45],"framework,":[46],"where":[47],"a":[48,53,57],"trainable":[49],"lightweight":[50],"projector":[51],"connects":[52],"speech":[54],"encoder":[55],"and":[56,126],"LLM.":[58],"Firstly,":[59],"we":[60,77,107],"assess":[61],"training":[62,98],"data":[63,93],"volume":[64],"requirements":[65],"to":[66],"match":[67],"Whisper-only":[68],"performance,":[69],"re-emphasizing":[70],"challenges":[72],"limited":[74],"data.":[75],"Secondly,":[76],"show":[78],"that":[79],"leveraging":[80],"mono-":[81],"or":[82],"multilingual":[83,101],"projectors":[84],"pretrained":[85],"on":[86,110,119],"languages":[88,125],"reduces":[89],"impact":[91],"scarcity,":[94],"especially":[95],"with":[96,105],"small":[97],"sets.":[99],"Using":[100],"(EuroLLM,":[103],"Salamandra)":[104],"whisper-large-v3-turbo,":[106],"evaluate":[108],"several":[111],"public":[112],"benchmarks,":[113],"providing":[114],"insights":[115],"future":[117],"research":[118],"optimizing":[120],"multilinguality.":[127]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2026-08-22T07:34:49.880490","created_date":"2025-10-24T00:00:00"}