{"id":"https://openalex.org/W4415133994","doi":"https://doi.org/10.21437/interspeech.2025-2260","title":"Overcoming Data Scarcity in Multi-Dialectal Arabic ASR via Whisper Fine-Tuning","display_name":"Overcoming Data Scarcity in Multi-Dialectal Arabic ASR via Whisper Fine-Tuning","publication_year":2025,"publication_date":"2025-08-17","ids":{"openalex":"https://openalex.org/W4415133994","doi":"https://doi.org/10.21437/interspeech.2025-2260"},"language":"en","primary_location":{"id":"doi:10.21437/interspeech.2025-2260","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2025-2260","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":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://research.rug.nl/en/publications/127d88f1-0d47-40eb-9061-ff1957d74144","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5119787792","display_name":"\u00d6mer Tarik \u00d6zyilmaz","orcid":"https://orcid.org/0009-0005-3911-5520"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"\u00d6mer Tarik \u00d6zyilmaz","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5090819693","display_name":"Matt Coler","orcid":"https://orcid.org/0000-0002-7631-5063"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Matt Coler","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5013519837","display_name":"Mat\u00edas Valdenegro-Toro","orcid":"https://orcid.org/0000-0001-5793-9498"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Matias Valdenegro-Toro","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":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.24607487,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1158","last_page":"1162"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9991000294685364,"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":0.9991000294685364,"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/T12031","display_name":"Speech and dialogue systems","score":0.994700014591217,"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":0.9934999942779541,"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/modern-standard-arabic","display_name":"Modern Standard Arabic","score":0.8478999733924866},{"id":"https://openalex.org/keywords/scarcity","display_name":"Scarcity","score":0.6658999919891357},{"id":"https://openalex.org/keywords/arabic","display_name":"Arabic","score":0.6496999859809875},{"id":"https://openalex.org/keywords/pooling","display_name":"Pooling","score":0.640999972820282},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.5758000016212463},{"id":"https://openalex.org/keywords/morpheme","display_name":"Morpheme","score":0.390500009059906}],"concepts":[{"id":"https://openalex.org/C2778243841","wikidata":"https://www.wikidata.org/wiki/Q56467","display_name":"Modern Standard Arabic","level":3,"score":0.8478999733924866},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6980999708175659},{"id":"https://openalex.org/C109747225","wikidata":"https://www.wikidata.org/wiki/Q815758","display_name":"Scarcity","level":2,"score":0.6658999919891357},{"id":"https://openalex.org/C96455323","wikidata":"https://www.wikidata.org/wiki/Q13955","display_name":"Arabic","level":2,"score":0.6496999859809875},{"id":"https://openalex.org/C70437156","wikidata":"https://www.wikidata.org/wiki/Q7228652","display_name":"Pooling","level":2,"score":0.640999972820282},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.6305000185966492},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.5758000016212463},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5130000114440918},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.4097999930381775},{"id":"https://openalex.org/C165297611","wikidata":"https://www.wikidata.org/wiki/Q43249","display_name":"Morpheme","level":2,"score":0.390500009059906},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.3709999918937683},{"id":"https://openalex.org/C134121241","wikidata":"https://www.wikidata.org/wiki/Q899301","display_name":"Yield (engineering)","level":2,"score":0.3504999876022339},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.33340001106262207},{"id":"https://openalex.org/C194051981","wikidata":"https://www.wikidata.org/wiki/Q1337691","display_name":"Economic shortage","level":3,"score":0.25369998812675476}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.21437/interspeech.2025-2260","is_oa":false,"landing_page_url":"https://doi.org/10.21437/interspeech.2025-2260","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:pure.rug.nl:openaire/127d88f1-0d47-40eb-9061-ff1957d74144","is_oa":true,"landing_page_url":"https://research.rug.nl/en/publications/127d88f1-0d47-40eb-9061-ff1957d74144","pdf_url":null,"source":{"id":"https://openalex.org/S4306400420","display_name":"University of Groningen research database (University of Groningen / Centre for Information Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I169381384","host_organization_name":"University of Groningen","host_organization_lineage":["https://openalex.org/I169381384"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"\u00d6zyilmaz, \u00d6, Coler, M & Valdenegro-Toro, M 2025, Overcoming Data Scarcity in Multi-Dialectal Arabic ASR via Whisper Fine-Tuning. in Proceedings of Interspeech 2025. ISCA, pp. 1158-1162, Interspeech 2025, Rotterdam, Netherlands, 17/08/2025. https://doi.org/10.21437/Interspeech.2025-2260","raw_type":"info:eu-repo/semantics/conferenceObject"},{"id":"pmh:oai:arXiv.org:2506.02627","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2506.02627","pdf_url":"https://arxiv.org/pdf/2506.02627","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"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":"text"},{"id":"pmh:oai:pure.rug.nl:openaire_cris_publications/127d88f1-0d47-40eb-9061-ff1957d74144","is_oa":true,"landing_page_url":"https://hdl.handle.net/11370/127d88f1-0d47-40eb-9061-ff1957d74144","pdf_url":null,"source":{"id":"https://openalex.org/S4306400420","display_name":"University of Groningen research database (University of Groningen / Centre for Information Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I169381384","host_organization_name":"University of Groningen","host_organization_lineage":["https://openalex.org/I169381384"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"\u00d6zyilmaz, \u00d6, Coler, M & Valdenegro-Toro, M 2025, Overcoming Data Scarcity in Multi-Dialectal Arabic ASR via Whisper Fine-Tuning. in Proceedings of Interspeech 2025. ISCA, pp. 1158-1162, Interspeech 2025, Rotterdam, Netherlands, 17/08/2025. https://doi.org/10.21437/Interspeech.2025-2260","raw_type":"info:eu-repo/semantics/conferenceObject"}],"best_oa_location":{"id":"pmh:oai:pure.rug.nl:openaire/127d88f1-0d47-40eb-9061-ff1957d74144","is_oa":true,"landing_page_url":"https://research.rug.nl/en/publications/127d88f1-0d47-40eb-9061-ff1957d74144","pdf_url":null,"source":{"id":"https://openalex.org/S4306400420","display_name":"University of Groningen research database (University of Groningen / Centre for Information Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I169381384","host_organization_name":"University of Groningen","host_organization_lineage":["https://openalex.org/I169381384"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"\u00d6zyilmaz, \u00d6, Coler, M & Valdenegro-Toro, M 2025, Overcoming Data Scarcity in Multi-Dialectal Arabic ASR via Whisper Fine-Tuning. in Proceedings of Interspeech 2025. ISCA, pp. 1158-1162, Interspeech 2025, Rotterdam, Netherlands, 17/08/2025. https://doi.org/10.21437/Interspeech.2025-2260","raw_type":"info:eu-repo/semantics/conferenceObject"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Although":[0],"commercial":[1],"Arabic":[2,11,29],"automatic":[3],"speech":[4],"recognition":[5],"(ASR)":[6],"systems":[7],"support":[8],"Modern":[9],"Standard":[10],"(MSA),":[12],"they":[13],"struggle":[14],"with":[15],"dialectal":[16,47,111],"speech.":[17,48],"We":[18,49,66],"investigate":[19],"the":[20,43],"effect":[21],"of":[22,56,71],"fine-tuning":[23,73],"OpenAI's":[24],"Whisper":[25],"on":[26,58],"five":[27],"major":[28],"dialects":[30],"(Gulf,":[31],"Levantine,":[32],"Iraqi,":[33],"Egyptian,":[34],"Maghrebi)":[35],"using":[36],"Mozilla":[37],"Common":[38],"Voice":[39],"for":[40,46,78],"MSA":[41,51,59,72,86,96],"and":[42,61,97],"MASC":[44],"dataset":[45],"evaluate":[50],"training":[52],"size":[53],"effects,":[54],"benefits":[55],"pre-training":[57,87],"data,":[60,112],"dialect-specific":[62,105],"versus":[63],"dialect-pooled":[64,100],"models.":[65,84],"find":[67],"that":[68,109],"small":[69],"amounts":[70],"data":[74,119],"yield":[75],"substantial":[76],"improvements":[77],"smaller":[79],"models,":[80],"matching":[81],"larger":[82],"non-fine-tuned":[83],"While":[85],"shows":[88],"minimal":[89],"benefit,":[90],"suggesting":[91],"limited":[92],"shared":[93],"features":[94],"between":[95],"dialects,":[98],"our":[99],"models":[101],"perform":[102],"comparably":[103],"to":[104],"ones.":[106],"This":[107],"indicates":[108],"pooling":[110],"when":[113],"properly":[114],"balanced,":[115],"can":[116],"help":[117],"address":[118],"scarcity":[120],"in":[121],"low-resource":[122],"ASR":[123],"without":[124],"significant":[125],"performance":[126],"loss.":[127]},"counts_by_year":[],"updated_date":"2026-08-01T09:00:35.917206","created_date":"2025-10-14T00:00:00"}