{"id":"https://openalex.org/W7167259513","doi":"https://doi.org/10.48550/arxiv.2607.02435","title":"MARVEL: Margin-Aware Robust von Mises-Fischer Expert Learning for Long-Tailed Out-of-Distribution Detection","display_name":"MARVEL: Margin-Aware Robust von Mises-Fischer Expert Learning for Long-Tailed Out-of-Distribution Detection","publication_year":2026,"publication_date":"2026-07-02","ids":{"openalex":"https://openalex.org/W7167259513","doi":"https://doi.org/10.48550/arxiv.2607.02435"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2607.02435","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.02435","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,"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":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2607.02435","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5140015032","display_name":"A. S. Anudeep","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Anudeep, A. S.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5029254671","display_name":"Vaanathi Sundaresan","orcid":"https://orcid.org/0000-0002-9451-4779"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sundaresan, Vaanathi","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":false,"primary_topic":{"id":"https://openalex.org/T11652","display_name":"Imbalanced Data Classification Techniques","score":0.2651999890804291,"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/T11652","display_name":"Imbalanced Data Classification Techniques","score":0.2651999890804291,"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/T13702","display_name":"Machine Learning in Healthcare","score":0.25270000100135803,"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/T11512","display_name":"Anomaly Detection Techniques and Applications","score":0.08259999752044678,"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/outlier","display_name":"Outlier","score":0.6269999742507935},{"id":"https://openalex.org/keywords/classifier","display_name":"Classifier (UML)","score":0.4788999855518341},{"id":"https://openalex.org/keywords/identification","display_name":"Identification (biology)","score":0.4636000096797943},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.4348999857902527},{"id":"https://openalex.org/keywords/safer","display_name":"SAFER","score":0.3815999925136566},{"id":"https://openalex.org/keywords/anomaly-detection","display_name":"Anomaly detection","score":0.37290000915527344}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6830999851226807},{"id":"https://openalex.org/C79337645","wikidata":"https://www.wikidata.org/wiki/Q779824","display_name":"Outlier","level":2,"score":0.6269999742507935},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.6201000213623047},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6197999715805054},{"id":"https://openalex.org/C95623464","wikidata":"https://www.wikidata.org/wiki/Q1096149","display_name":"Classifier (UML)","level":2,"score":0.4788999855518341},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.4636000096797943},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.4348999857902527},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.387800008058548},{"id":"https://openalex.org/C2776654903","wikidata":"https://www.wikidata.org/wiki/Q2601463","display_name":"SAFER","level":2,"score":0.3815999925136566},{"id":"https://openalex.org/C739882","wikidata":"https://www.wikidata.org/wiki/Q3560506","display_name":"Anomaly detection","level":2,"score":0.37290000915527344},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.34450000524520874},{"id":"https://openalex.org/C2777548347","wikidata":"https://www.wikidata.org/wiki/Q5456937","display_name":"Flagging","level":2,"score":0.3239000141620636},{"id":"https://openalex.org/C58328972","wikidata":"https://www.wikidata.org/wiki/Q184609","display_name":"Expert system","level":2,"score":0.31679999828338623},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.3057999908924103},{"id":"https://openalex.org/C34736171","wikidata":"https://www.wikidata.org/wiki/Q918333","display_name":"Preprocessor","level":2,"score":0.26899999380111694}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2607.02435","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.02435","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,"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":"doi:10.48550/arxiv.2607.02435","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.02435","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,"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":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/16","display_name":"Peace, Justice and strong institutions","score":0.7318583130836487}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"For":[0],"clinical":[1],"deployment,":[2],"it":[3],"is":[4,221],"essential":[5],"that":[6,194],"automated":[7],"diagnostic":[8],"systems":[9],"remain":[10],"reliable":[11,203],"when":[12],"confronted":[13],"with":[14,26,116],"previously":[15],"unseen":[16],"cases,":[17],"yet":[18],"deep":[19],"models":[20],"routinely":[21],"misclassify":[22],"out-of-distribution":[23],"(OOD)":[24],"inputs":[25],"high":[27],"confidence,":[28],"underscoring":[29],"the":[30,50,75,196],"need":[31],"for":[32,208],"more":[33],"robust":[34],"OOD":[35,57,63,71,95,161],"detection":[36,58],"methods.":[37],"Although":[38],"substantial":[39],"effort":[40],"has":[41],"been":[42],"devoted":[43],"to":[44,123,142,153,210],"improving":[45],"model":[46],"robustness,":[47],"most":[48],"of":[49,111,119,139,180,198,205],"existing":[51],"literature":[52],"assumes":[53],"balanced":[54],"datasets,":[55],"evaluates":[56],"on":[59,83,164],"coarse":[60],"or":[61,65],"non-clinical":[62],"sources,":[64],"lacks":[66],"comprehensive":[67,192],"assessment":[68],"across":[69,91],"diverse":[70,84],"scenarios.":[72],"To":[73],"address":[74],"gaps,":[76],"we":[77],"propose":[78],"a":[79,92,104,127],"novel":[80],"methodology":[81],"trained":[82,151],"and":[85,89,146,167,183],"imbalanced":[86],"medical":[87],"datasets":[88,169],"evaluated":[90],"clinically":[93],"reflective":[94],"spectrum.":[96],"Our":[97,219],"framework":[98,129],"comprises":[99],"three":[100],"key":[101],"components:":[102],"(1)":[103],"Nonlinear":[105],"von":[106],"Mises-Fisher":[107],"(NvMF)":[108],"classifier":[109],"capable":[110],"learning":[112],"non-linear":[113],"decision":[114],"boundaries,":[115],"theoretical":[117],"proof":[118],"its":[120],"asymptotic":[121],"connection":[122],"cosine":[124],"classifiers;":[125],"(2)":[126],"multi-expert":[128],"in":[130,136,216],"which":[131],"margin-aware":[132],"NvMF":[133],"classifiers":[134],"specialise":[135],"different":[137],"regions":[138],"label":[140],"distribution":[141],"better":[143],"handle":[144],"imbalance;":[145],"(3)":[147],"an":[148],"outlier":[149,157],"expert":[150],"explicitly":[152],"distinguish":[154],"inlier":[155],"from":[156],"data,":[158],"thereby":[159],"strengthening":[160],"detection.":[162],"Evaluation":[163],"RFMiD,":[165],"ISIC2019,":[166],"NCTCRC":[168],"demonstrates":[170],"consistent":[171],"improvements":[172],"over":[173],"state-of-the-art":[174],"methods,":[175],"achieving":[176],"mean":[177],"FPR95":[178],"reductions":[179],"8.45%,":[181],"13.02%,":[182],"36.90%":[184],"respectively.":[185],"These":[186],"gains":[187],"are":[188],"further":[189],"supported":[190],"by":[191],"ablations":[193],"validated":[195],"contributions":[197],"each":[199],"component.":[200],"This":[201],"enables":[202],"identification":[204],"unfamiliar":[206],"cases":[207],"deferral":[209],"clinicians,":[211],"supporting":[212],"safer":[213],"AI-assisted":[214],"diagnosis":[215],"real-world":[217],"workflows.":[218],"code":[220],"available":[222],"at":[223],"https://github.com/redboxup/MARVEL.":[224]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-07-04T00:00:00"}