{"id":"https://openalex.org/W6910691227","doi":"https://doi.org/10.48550/arxiv.2506.17499","title":"Episode-specific Fine-tuning for Metric-based Few-shot Learners with Optimization-based Training","display_name":"Episode-specific Fine-tuning for Metric-based Few-shot Learners with Optimization-based Training","publication_year":2025,"publication_date":"2025-06-20","ids":{"openalex":"https://openalex.org/W6910691227","doi":"https://doi.org/10.48550/arxiv.2506.17499"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2506.17499","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2506.17499","pdf_url":"https://arxiv.org/pdf/2506.17499","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,"listed_in":[],"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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2506.17499","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5113419002","display_name":"Xuanyu Zhuang","orcid":"https://orcid.org/0009-0008-1003-741X","observed_orcids":["https://orcid.org/0009-0008-1003-741X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhuang, Xuanyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5063097936","display_name":"Geoffroy G. Peeters","orcid":"https://orcid.org/0000-0001-5255-3019","observed_orcids":["https://orcid.org/0000-0001-5255-3019"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Peeters, Geoffroy","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5055423112","display_name":"Ga\u00ebl Richard A","orcid":"https://orcid.org/0000-0002-4960-0010","observed_orcids":["https://orcid.org/0000-0002-4960-0010"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Richard, Ga\u00ebl","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":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":true,"primary_topic":{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.9897000193595886,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.9897000193595886,"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/T11309","display_name":"Music and Audio Processing","score":0.010099999606609344,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/T10028","display_name":"Topic Modeling","score":0.0,"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/few-shot-classification","display_name":"few-shot classification","score":0.765999972820282},{"id":"https://openalex.org/keywords/overfitting-reduction","display_name":"overfitting reduction","score":0.6620000004768372},{"id":"https://openalex.org/keywords/audio-classification","display_name":"audio classification","score":0.7770000100135803},{"id":"https://openalex.org/keywords/attention-models","display_name":"attention-based models","score":0.5619999766349792}],"concepts":[{"id":"https://openalex.org/C22019652","wikidata":"https://www.wikidata.org/wiki/Q331309","display_name":"Overfitting","level":3,"score":0.8840000033378601},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6870999932289124},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.656000018119812},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6251999735832214},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.6219000220298767},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5508000254631042},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.5400000214576721},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5375000238418579},{"id":"https://openalex.org/C103278499","wikidata":"https://www.wikidata.org/wiki/Q254465","display_name":"Similarity (geometry)","level":3,"score":0.5095999836921692},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.47749999165534973},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.4275999963283539},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3513999879360199},{"id":"https://openalex.org/C60798267","wikidata":"https://www.wikidata.org/wiki/Q1226939","display_name":"Division (mathematics)","level":2,"score":0.3407000005245209},{"id":"https://openalex.org/C2778572836","wikidata":"https://www.wikidata.org/wiki/Q380933","display_name":"Space (punctuation)","level":2,"score":0.3353999853134155},{"id":"https://openalex.org/C2780586882","wikidata":"https://www.wikidata.org/wiki/Q7520643","display_name":"Simple (philosophy)","level":2,"score":0.3192000091075897},{"id":"https://openalex.org/C58489278","wikidata":"https://www.wikidata.org/wiki/Q1172284","display_name":"Data set","level":2,"score":0.3066999912261963},{"id":"https://openalex.org/C198043062","wikidata":"https://www.wikidata.org/wiki/Q180953","display_name":"Metric space","level":2,"score":0.29919999837875366},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.28949999809265137},{"id":"https://openalex.org/C143724316","wikidata":"https://www.wikidata.org/wiki/Q312468","display_name":"Series (stratigraphy)","level":2,"score":0.27250000834465027},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.2538999915122986},{"id":"https://openalex.org/C27158222","wikidata":"https://www.wikidata.org/wiki/Q5532422","display_name":"Generalizability theory","level":2,"score":0.25380000472068787}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2506.17499","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2506.17499","pdf_url":"https://arxiv.org/pdf/2506.17499","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,"listed_in":[],"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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2506.17499","doi":"https://doi.org/10.48550/arxiv.2506.17499","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2506.17499","pdf_url":null,"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,"listed_in":[],"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2506.17499","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2506.17499","pdf_url":"https://arxiv.org/pdf/2506.17499","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,"listed_in":[],"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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[{"id":"https://metadata.un.org/sdg/4","score":0.7253287434577942,"display_name":"Quality education"}],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"In":[0],"few-shot":[1],"classification":[2,20],"tasks":[3],"(so-called":[4],"episodes),":[5],"a":[6,38,83,143],"small":[7],"set":[8,124],"of":[9,21,85,137,146,173],"labeled":[10,48],"support":[11,35,49,123,192],"samples":[12,50,193],"is":[13],"provided":[14],"during":[15,194],"inference":[16,195],"to":[17,63,71,125,159,186,189],"aid":[18],"the":[19,67,72,75,121,133,161,170,184,190],"unlabeled":[22],"query":[23,33],"samples.":[24],"Metric-based":[25],"models":[26,180,236],"typically":[27],"operate":[28],"by":[29,43],"computing":[30],"similarities":[31],"between":[32],"and":[34,65,101,109,176,219,241],"embeddings":[36],"within":[37,164],"learned":[39],"metric":[40,68],"space,":[41],"followed":[42],"nearest-neighbor":[44],"classification.":[45],"However,":[46],"these":[47],"are":[51,54,181],"often":[52],"underutilized--they":[53],"only":[55],"used":[56],"for":[57,93,129,232,238],"similarity":[58],"comparison,":[59],"despite":[60],"their":[61],"potential":[62],"fine-tune":[64],"adapt":[66,188],"space":[69],"itself":[70],"classes":[73],"in":[74,139],"current":[76],"episode.":[77],"To":[78,153],"address":[79],"this,":[80,155],"we":[81,156],"propose":[82,158],"series":[84],"simple":[86],"yet":[87],"effective":[88],"episode-specific,":[89],"during-inference":[90],"fine-tuning":[91,127,151,175],"methods":[92,115],"metric-based":[94,162,179,235],"models,":[95],"including":[96],"Rotational":[97],"Division":[98,106,111],"Fine-Tuning":[99,107,112],"(RDFT)":[100],"its":[102],"two":[103],"variants,":[104],"Iterative":[105],"(IDFT)":[108],"Augmented":[110],"(ADFT).":[113],"These":[114],"construct":[116],"pseudo":[117],"support-query":[118],"pairs":[119],"from":[120,207],"given":[122],"enable":[126],"even":[128],"non-parametric":[130],"models.":[131],"Nevertheless,":[132],"severely":[134],"limited":[135,191],"amount":[136],"data":[138],"each":[140],"task":[141],"poses":[142],"substantial":[144],"risk":[145],"overfitting":[147],"when":[148],"applying":[149],"such":[150],"strategies.":[152],"mitigate":[154],"further":[157],"train":[160],"model":[163],"an":[165],"optimization-based":[166,177],"meta-learning":[167],"framework.":[168],"With":[169],"combined":[171],"efforts":[172],"episode-specific":[174],"meta-training,":[178],"equipped":[182],"with":[183],"ability":[185],"rapidly":[187],"while":[196],"avoiding":[197],"overfitting.":[198],"We":[199],"validate":[200],"our":[201,227],"approach":[202,228],"on":[203],"three":[204],"audio":[205,246],"datasets":[206],"diverse":[208],"domains,":[209],"namely":[210],"ESC-50":[211],"(environmental":[212],"sounds),":[213],"Speech":[214],"Commands":[215],"V2":[216],"(spoken":[217],"keywords),":[218],"Medley-solos-DB":[220],"(musical":[221],"instrument).":[222],"Experimental":[223],"results":[224],"demonstrate":[225],"that":[226],"consistently":[229],"improves":[230],"performance":[231],"all":[233],"evaluated":[234],"(especially":[237],"attention-based":[239],"models)":[240],"generalizes":[242],"well":[243],"across":[244],"different":[245],"domains.":[247]},"counts_by_year":[],"updated_date":"2026-10-14T16:39:01.820549","created_date":"2025-10-10T00:00:00"}