{"id":"https://openalex.org/W7138916367","doi":"https://doi.org/10.48550/arxiv.2603.17173","title":"Generalist Multimodal LLMs Gain Biometric Expertise via Human Salience","display_name":"Generalist Multimodal LLMs Gain Biometric Expertise via Human Salience","publication_year":2026,"publication_date":"2026-03-17","ids":{"openalex":"https://openalex.org/W7138916367","doi":"https://doi.org/10.48550/arxiv.2603.17173"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.17173","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.17173","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2603.17173","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5020327450","display_name":"Jacob Piland","orcid":"https://orcid.org/0009-0000-6140-7734"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Piland, Jacob","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5120732948","display_name":"Byron Dowling","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dowling, Byron","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103104238","display_name":"Christopher Sweet","orcid":"https://orcid.org/0000-0001-8214-7177"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sweet, Christopher","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5129878814","display_name":"Adam Czajka","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Czajka, Adam","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":4,"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/T10828","display_name":"Biometric Identification and Security","score":0.5954999923706055,"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"}},"topics":[{"id":"https://openalex.org/T10828","display_name":"Biometric Identification and Security","score":0.5954999923706055,"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/T11448","display_name":"Face recognition and analysis","score":0.16580000519752502,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.05889999866485596,"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/biometrics","display_name":"Biometrics","score":0.7635999917984009},{"id":"https://openalex.org/keywords/iris-recognition","display_name":"Iris recognition","score":0.5708000063896179},{"id":"https://openalex.org/keywords/salience","display_name":"Salience (neuroscience)","score":0.5568000078201294},{"id":"https://openalex.org/keywords/cluster-analysis","display_name":"Cluster analysis","score":0.46389999985694885},{"id":"https://openalex.org/keywords/pointer","display_name":"Pointer (user interface)","score":0.35339999198913574},{"id":"https://openalex.org/keywords/threat-model","display_name":"Threat model","score":0.3483999967575073},{"id":"https://openalex.org/keywords/ambiguity","display_name":"Ambiguity","score":0.34200000762939453},{"id":"https://openalex.org/keywords/profiling","display_name":"Profiling (computer programming)","score":0.3330000042915344},{"id":"https://openalex.org/keywords/robustness","display_name":"Robustness (evolution)","score":0.32440000772476196}],"concepts":[{"id":"https://openalex.org/C184297639","wikidata":"https://www.wikidata.org/wiki/Q177765","display_name":"Biometrics","level":2,"score":0.7635999917984009},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7412999868392944},{"id":"https://openalex.org/C112356035","wikidata":"https://www.wikidata.org/wiki/Q1672722","display_name":"Iris recognition","level":3,"score":0.5708000063896179},{"id":"https://openalex.org/C108154423","wikidata":"https://www.wikidata.org/wiki/Q1469792","display_name":"Salience (neuroscience)","level":2,"score":0.5568000078201294},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5329999923706055},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.48260000348091125},{"id":"https://openalex.org/C73555534","wikidata":"https://www.wikidata.org/wiki/Q622825","display_name":"Cluster analysis","level":2,"score":0.46389999985694885},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4535999894142151},{"id":"https://openalex.org/C150202949","wikidata":"https://www.wikidata.org/wiki/Q107602","display_name":"Pointer (user interface)","level":2,"score":0.35339999198913574},{"id":"https://openalex.org/C140547941","wikidata":"https://www.wikidata.org/wiki/Q7797194","display_name":"Threat model","level":2,"score":0.3483999967575073},{"id":"https://openalex.org/C2780522230","wikidata":"https://www.wikidata.org/wiki/Q1140419","display_name":"Ambiguity","level":2,"score":0.34200000762939453},{"id":"https://openalex.org/C187191949","wikidata":"https://www.wikidata.org/wiki/Q1138496","display_name":"Profiling (computer programming)","level":2,"score":0.3330000042915344},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.32440000772476196},{"id":"https://openalex.org/C101738243","wikidata":"https://www.wikidata.org/wiki/Q786435","display_name":"Autoencoder","level":3,"score":0.3237999975681305},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3181999921798706},{"id":"https://openalex.org/C123201435","wikidata":"https://www.wikidata.org/wiki/Q456632","display_name":"Information privacy","level":2,"score":0.31209999322891235},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.3107999861240387},{"id":"https://openalex.org/C108827166","wikidata":"https://www.wikidata.org/wiki/Q175975","display_name":"Internet privacy","level":1,"score":0.30329999327659607},{"id":"https://openalex.org/C2779965156","wikidata":"https://www.wikidata.org/wiki/Q5227350","display_name":"Data sharing","level":3,"score":0.30219998955726624},{"id":"https://openalex.org/C2779267917","wikidata":"https://www.wikidata.org/wiki/Q170028","display_name":"Deception","level":2,"score":0.30149999260902405},{"id":"https://openalex.org/C62230096","wikidata":"https://www.wikidata.org/wiki/Q275969","display_name":"Crowdsourcing","level":2,"score":0.2985000014305115},{"id":"https://openalex.org/C23224414","wikidata":"https://www.wikidata.org/wiki/Q176769","display_name":"Hidden Markov model","level":2,"score":0.2921999990940094},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.29159998893737793},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.28870001435279846},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.2851000130176544},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.27790001034736633},{"id":"https://openalex.org/C2779089604","wikidata":"https://www.wikidata.org/wiki/Q7169333","display_name":"Permission","level":2,"score":0.2773999869823456},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.26919999718666077},{"id":"https://openalex.org/C2779623668","wikidata":"https://www.wikidata.org/wiki/Q7652842","display_name":"SwIPe","level":2,"score":0.2635999917984009},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.2538999915122986},{"id":"https://openalex.org/C43091099","wikidata":"https://www.wikidata.org/wiki/Q1067788","display_name":"Through-the-lens metering","level":3,"score":0.25220000743865967},{"id":"https://openalex.org/C79974875","wikidata":"https://www.wikidata.org/wiki/Q483639","display_name":"Cloud computing","level":2,"score":0.25209999084472656}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.17173","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.17173","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2603.17173","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.17173","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"article"},"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":{"Iris":[0],"presentation":[1],"attack":[2,55,120,136,151,170],"detection":[3],"(PAD)":[4],"is":[5,25,40],"critical":[6],"for":[7,127,225],"secure":[8],"biometric":[9,44,91],"deployments,":[10],"yet":[11,31],"developing":[12],"specialized":[13,195],"models":[14,71,155,181],"faces":[15],"significant":[16],"practical":[17],"barriers:":[18],"collecting":[19,28],"data":[20,45,92],"representing":[21],"future":[22],"unknown":[23],"attacks":[24],"impossible,":[26],"and":[27,201],"diverse-enough":[29],"data,":[30],"still":[32],"limited":[33],"in":[34,63,114],"terms":[35],"of":[36,53,100,164],"its":[37],"predictive":[38],"power,":[39],"expensive.":[41],"Additionally,":[42],"sharing":[43],"raises":[46],"privacy":[47,86,219],"concerns.":[48],"Due":[49],"to":[50,93,105,156],"rapid":[51],"emergence":[52],"new":[54],"vectors":[56],"demanding":[57],"adaptable":[58],"solutions,":[59],"we":[60,108,138,185],"thus":[61],"investigate":[62],"this":[64,128],"paper":[65],"whether":[66],"general-purpose":[67],"multimodal":[68],"large":[69],"language":[70],"(MLLMs)":[72],"can":[73],"perform":[74],"iris":[75,119,166,226],"PAD":[76],"when":[77],"augmented":[78],"with":[79,189],"human":[80,144,202],"expert":[81],"knowledge,":[82],"operating":[83],"under":[84],"strict":[85],"constraints":[87,220],"that":[88,110,140,187,214],"prohibit":[89],"sending":[90],"public":[94],"cloud":[95],"MLLM":[96],"services.":[97],"Through":[98],"analysis":[99],"vision":[101,112],"encoder":[102],"embeddings":[103],"applied":[104],"our":[106],"dataset,":[107],"demonstrate":[109],"pre-trained":[111],"transformers":[113],"MLLMs":[115,215],"inherently":[116],"cluster":[117],"many":[118],"types":[121],"despite":[122],"never":[123],"being":[124],"explicitly":[125],"trained":[126],"task.":[129],"However,":[130],"where":[131],"clustering":[132],"shows":[133],"overlap":[134],"between":[135],"classes,":[137],"find":[139],"structured":[141],"prompts":[142,191],"incorporating":[143],"salience":[145],"(verbal":[146],"descriptions":[147],"from":[148],"subjects":[149],"identifying":[150],"indicators)":[152],"enable":[153],"these":[154],"resolve":[157],"ambiguities.":[158],"Testing":[159],"on":[160],"an":[161],"IRB-restricted":[162],"dataset":[163],"224":[165],"images":[167],"spanning":[168],"seven":[169],"types,":[171],"using":[172],"only":[173],"university-approved":[174],"services":[175],"(Gemini":[176],"2.5":[177],"Pro)":[178],"or":[179],"locally-hosted":[180],"(e.g.,":[182],"Llama":[183,207],"3.2-Vision),":[184],"show":[186],"Gemini":[188],"expert-informed":[190],"outperforms":[192],"both":[193],"a":[194,222],"convolutional":[196],"neural":[197],"networks":[198],"(CNN)-based":[199],"baseline":[200],"examiners,":[203],"while":[204],"the":[205],"locally-deployable":[206],"achieves":[208],"near-human":[209],"performance.":[210],"Our":[211],"results":[212],"establish":[213],"deployable":[216],"within":[217],"institutional":[218],"offer":[221],"viable":[223],"path":[224],"PAD.":[227]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-03-20T00:00:00"}