{"id":"https://openalex.org/W7164346347","doi":"https://doi.org/10.48550/arxiv.2606.12023","title":"ViT-FREE: Efficient Face Recognition via Early Exiting and Synthetic Adaptation","display_name":"ViT-FREE: Efficient Face Recognition via Early Exiting and Synthetic Adaptation","publication_year":2026,"publication_date":"2026-06-10","ids":{"openalex":"https://openalex.org/W7164346347","doi":"https://doi.org/10.48550/arxiv.2606.12023"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.12023","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.12023","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":null,"license_id":null,"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.2606.12023","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5106356516","display_name":"Tahar Chettaoui","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chettaoui, Tahar","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5115792474","display_name":"Guray Ozgur","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ozgur, Guray","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138402106","display_name":"Eduarda Caldeira","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Caldeira, Eduarda","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138479360","display_name":"Naser Damer","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Damer, Naser","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5028035125","display_name":"Fadi Boutros","orcid":"https://orcid.org/0000-0003-4516-9128"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Boutros, Fadi","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/T11448","display_name":"Face recognition and analysis","score":0.9860000014305115,"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"}},"topics":[{"id":"https://openalex.org/T11448","display_name":"Face recognition and analysis","score":0.9860000014305115,"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/T10057","display_name":"Face and Expression Recognition","score":0.002099999925121665,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.002099999925121665,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.6450999975204468},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6065000295639038},{"id":"https://openalex.org/keywords/discriminative-model","display_name":"Discriminative model","score":0.6011000275611877},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.5174999833106995},{"id":"https://openalex.org/keywords/software-deployment","display_name":"Software deployment","score":0.5098000168800354},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.49239999055862427},{"id":"https://openalex.org/keywords/boosting","display_name":"Boosting (machine learning)","score":0.45579999685287476},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.40950000286102295},{"id":"https://openalex.org/keywords/curse-of-dimensionality","display_name":"Curse of dimensionality","score":0.39750000834465027}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7864999771118164},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.6450999975204468},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6080999970436096},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6065000295639038},{"id":"https://openalex.org/C97931131","wikidata":"https://www.wikidata.org/wiki/Q5282087","display_name":"Discriminative model","level":2,"score":0.6011000275611877},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.5174999833106995},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.5098000168800354},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.49239999055862427},{"id":"https://openalex.org/C46686674","wikidata":"https://www.wikidata.org/wiki/Q466303","display_name":"Boosting (machine learning)","level":2,"score":0.45579999685287476},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.40950000286102295},{"id":"https://openalex.org/C111030470","wikidata":"https://www.wikidata.org/wiki/Q1430460","display_name":"Curse of dimensionality","level":2,"score":0.39750000834465027},{"id":"https://openalex.org/C175291020","wikidata":"https://www.wikidata.org/wiki/Q1156822","display_name":"Offset (computer science)","level":2,"score":0.3774000108242035},{"id":"https://openalex.org/C2778712577","wikidata":"https://www.wikidata.org/wiki/Q3505966","display_name":"Retraining","level":2,"score":0.3659999966621399},{"id":"https://openalex.org/C31510193","wikidata":"https://www.wikidata.org/wiki/Q1192553","display_name":"Facial recognition system","level":3,"score":0.3635999858379364},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.3619000017642975},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.3537999987602234},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.3395000100135803},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3382999897003174},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.313400000333786},{"id":"https://openalex.org/C52622490","wikidata":"https://www.wikidata.org/wiki/Q1026626","display_name":"Feature extraction","level":2,"score":0.310699999332428},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.30390000343322754},{"id":"https://openalex.org/C2779304628","wikidata":"https://www.wikidata.org/wiki/Q3503480","display_name":"Face (sociological concept)","level":2,"score":0.29989999532699585},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.28049999475479126},{"id":"https://openalex.org/C2780598303","wikidata":"https://www.wikidata.org/wiki/Q65921492","display_name":"Flexibility (engineering)","level":2,"score":0.27410000562667847},{"id":"https://openalex.org/C70518039","wikidata":"https://www.wikidata.org/wiki/Q16000077","display_name":"Dimensionality reduction","level":2,"score":0.2612999975681305},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.26080000400543213},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.2531999945640564},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.25119999051094055},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.25110000371932983},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.25049999356269836}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.12023","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.12023","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.12023","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.12023","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/10","score":0.6944173574447632,"display_name":"Reduced inequalities"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Vision":[0],"Transformers":[1],"(ViTs)":[2],"have":[3],"gained":[4],"significant":[5],"attention":[6,102,128],"in":[7,46,188],"computer":[8],"vision":[9],"and":[10,37,90,101,115,127,137,235],"shown":[11],"strong":[12],"potential":[13],"for":[14,32,56,141],"face":[15,77],"recognition":[16],"(FR).":[17],"However,":[18],"their":[19],"high":[20,109],"computational":[21],"cost":[22],"makes":[23],"deployment":[24],"on":[25,147,191],"resource-constrained":[26],"devices":[27],"challenging,":[28],"motivating":[29],"the":[30,61,87,119,154,208,218,225,232],"need":[31],"methods":[33],"that":[34,75,98,131,163,205],"balance":[35],"efficiency":[36,233],"accuracy.":[38],"In":[39],"this":[40],"work,":[41],"we":[42,69,96,151,197],"investigate":[43],"early":[44,142],"exiting":[45,172],"pretrained":[47],"ViTs":[48],"as":[49,194],"a":[50,72,167,179,185,200,212],"simple":[51],"yet":[52],"effective":[53],"training-free":[54],"strategy":[55,204],"efficient":[57],"FR":[58,149],"inference.":[59],"Leveraging":[60],"uniform":[62],"feature":[63,125],"dimensionality":[64],"across":[65,106,157],"transformer":[66,219],"encoder":[67],"blocks,":[68],"introduce":[70],"ViT-FREE,":[71],"multi-exit":[73],"framework":[74],"enables":[76],"verification":[78,189],"directly":[79],"from":[80],"intermediate":[81,132],"representations":[82,139],"without":[83],"modifying":[84],"or":[85],"retraining":[86],"backbone":[88,220],"model,":[89],"thus,":[91],"reducing":[92],"inference":[93],"cost.":[94],"Empirically,":[95],"show":[97],"patch":[99],"embeddings":[100],"maps":[103],"evolve":[104],"progressively":[105],"depth,":[107],"exhibiting":[108],"similarity":[110],"between":[111],"consecutive":[112],"ViT":[113],"blocks":[114],"increasing":[116],"alignment":[117],"with":[118,171],"final":[120],"representation.":[121],"This":[122,222],"indicates":[123],"gradual":[124],"refinement":[126],"convergence,":[129],"suggesting":[130],"layers":[133,210],"already":[134],"provide":[135],"stable":[136],"discriminative":[138],"suitable":[140],"exiting.":[143],"Through":[144],"extensive":[145],"experiments":[146],"multiple":[148],"benchmarks,":[150],"systematically":[152],"analyze":[153],"accuracy-efficiency":[155],"trade-off":[156],"exit":[158],"depths.":[159],"Our":[160],"results":[161],"demonstrate":[162],"later":[164],"exits":[165,229,238],"achieve":[166],"highly":[168],"favorable":[169],"balance,":[170],"at":[173],"layer":[174],"10":[175],"yielding":[176],"up":[177],"to":[178],"20%":[180],"speedup":[181],"while":[182,216,230],"incurring":[183],"only":[184,207],"1.5":[186],"drop":[187],"performance":[190,226],"benchmarks":[192],"such":[193],"IJB-C.":[195],"Also,":[196],"propose":[198],"ViT-FREE_FT,":[199],"lightweight":[201],"exit-specific":[202],"fine-tuning":[203],"adapts":[206],"projection":[209],"using":[211],"small":[213],"synthetic":[214],"dataset":[215],"keeping":[217],"frozen.":[221],"approach":[223],"improves":[224],"of":[227],"shallow":[228],"preserving":[231],"benefits":[234],"leaving":[236],"deeper":[237],"largely":[239],"unaffected.":[240]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-12T00:00:00"}