{"id":"https://openalex.org/W7138179023","doi":"https://doi.org/10.48550/arxiv.2603.13374","title":"Geometry-Aware Semantic Reasoning for Training Free Video Anomaly Detection","display_name":"Geometry-Aware Semantic Reasoning for Training Free Video Anomaly Detection","publication_year":2026,"publication_date":"2026-03-10","ids":{"openalex":"https://openalex.org/W7138179023","doi":"https://doi.org/10.48550/arxiv.2603.13374"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.13374","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.13374","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":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.2603.13374","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5129740522","display_name":"Ali Zia","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zia, Ali","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129741837","display_name":"Usman Ali","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ali, Usman","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5035818012","display_name":"Muhammad Umer Ramzan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ramzan, Muhammad Umer","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129736032","display_name":"Hamza Abid","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Abid, Hamza","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129667064","display_name":"Abdul Rehman","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rehman, Abdul","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5129646984","display_name":"Wei Xiang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiang, Wei","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/T11512","display_name":"Anomaly Detection Techniques and Applications","score":0.9681000113487244,"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/T11512","display_name":"Anomaly Detection Techniques and Applications","score":0.9681000113487244,"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/T10812","display_name":"Human Pose and Action Recognition","score":0.017500000074505806,"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/T11439","display_name":"Video Analysis and Summarization","score":0.0017000000225380063,"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/anomaly-detection","display_name":"Anomaly detection","score":0.7419999837875366},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5960000157356262},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.5561000108718872},{"id":"https://openalex.org/keywords/mahalanobis-distance","display_name":"Mahalanobis distance","score":0.49129998683929443},{"id":"https://openalex.org/keywords/question-answering","display_name":"Question answering","score":0.4625999927520752},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.4602000117301941},{"id":"https://openalex.org/keywords/anomaly","display_name":"Anomaly (physics)","score":0.4564000070095062},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.4277999997138977},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.40700000524520874}],"concepts":[{"id":"https://openalex.org/C739882","wikidata":"https://www.wikidata.org/wiki/Q3560506","display_name":"Anomaly detection","level":2,"score":0.7419999837875366},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7221999764442444},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6560999751091003},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5960000157356262},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.5561000108718872},{"id":"https://openalex.org/C1921717","wikidata":"https://www.wikidata.org/wiki/Q1334846","display_name":"Mahalanobis distance","level":2,"score":0.49129998683929443},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.4625999927520752},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.4602000117301941},{"id":"https://openalex.org/C12997251","wikidata":"https://www.wikidata.org/wiki/Q567560","display_name":"Anomaly (physics)","level":2,"score":0.4564000070095062},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.4277999997138977},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.40700000524520874},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.38609999418258667},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.3711000084877014},{"id":"https://openalex.org/C165838908","wikidata":"https://www.wikidata.org/wiki/Q736777","display_name":"Calibration","level":2,"score":0.36719998717308044},{"id":"https://openalex.org/C2781122975","wikidata":"https://www.wikidata.org/wiki/Q16928266","display_name":"Semantic feature","level":2,"score":0.3653999865055084},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.353300005197525},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.35019999742507935},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.3472000062465668},{"id":"https://openalex.org/C83665646","wikidata":"https://www.wikidata.org/wiki/Q42139305","display_name":"Feature vector","level":2,"score":0.33719998598098755},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.33149999380111694},{"id":"https://openalex.org/C103278499","wikidata":"https://www.wikidata.org/wiki/Q254465","display_name":"Similarity (geometry)","level":3,"score":0.325300008058548},{"id":"https://openalex.org/C52622490","wikidata":"https://www.wikidata.org/wiki/Q1026626","display_name":"Feature extraction","level":2,"score":0.3174000084400177},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3118000030517578},{"id":"https://openalex.org/C130318100","wikidata":"https://www.wikidata.org/wiki/Q2268914","display_name":"Semantic similarity","level":2,"score":0.30070000886917114},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.3000999987125397},{"id":"https://openalex.org/C13672336","wikidata":"https://www.wikidata.org/wiki/Q3460803","display_name":"Bag-of-words model","level":2,"score":0.2822999954223633},{"id":"https://openalex.org/C94124525","wikidata":"https://www.wikidata.org/wiki/Q912550","display_name":"Categorization","level":2,"score":0.25110000371932983}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.13374","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.13374","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":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.2603.13374","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.13374","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":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/16","score":0.6391389966011047,"display_name":"Peace, Justice and strong institutions"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Training-free":[0],"video":[1],"anomaly":[2,30,70,97],"detection":[3,71],"(VAD)":[4],"has":[5],"recently":[6],"emerged":[7],"as":[8,72],"a":[9,28,59,106,143,189],"scalable":[10],"alternative":[11,193],"to":[12,35,43,90,194],"supervised":[13],"approaches,":[14],"yet":[15],"existing":[16],"methods":[17],"largely":[18],"rely":[19],"on":[20,164,171],"static":[21,195],"prompting":[22],"and":[23,46,95,166,169,174,184,191],"geometry-agnostic":[24],"feature":[25,79],"fusion.":[26],"As":[27],"result,":[29],"inference":[31,75],"is":[32,115],"often":[33],"reduced":[34],"shallow":[36],"similarity":[37],"matching":[38,197],"over":[39,105,157],"Euclidean":[40,196],"embeddings,":[41],"leading":[42],"unstable":[44],"predictions":[45,137],"limited":[47],"interpretability,":[48],"especially":[49],"in":[50,138,198],"complex":[51],"or":[52],"hierarchically":[53],"structured":[54],"scenes.":[55],"We":[56],"introduce":[57],"MM-VAD,":[58],"geometry-aware":[60,182],"semantic":[61,136,186],"reasoning":[62],"framework":[63],"for":[64],"training":[65],"free":[66],"VAD":[67],"that":[68,147,181],"reframes":[69],"adaptive":[73,101,185],"test-time":[74],"rather":[76],"than":[77],"fixed":[78],"comparison.":[80],"Our":[81,178],"approach":[82],"projects":[83],"caption-derived":[84],"scene":[85],"representations":[86],"into":[87],"hyperbolic":[88],"space":[89],"better":[91],"preserve":[92],"hierarchical":[93],"structure":[94],"performs":[96],"assessment":[98],"through":[99],"an":[100,121],"question":[102],"answering":[103],"process":[104],"frozen":[107],"large":[108],"language":[109],"model.":[110],"A":[111],"lightweight,":[112],"learnable":[113],"prompt":[114],"optimised":[116],"at":[117],"test":[118],"time":[119],"using":[120],"unsupervised":[122],"confidence-sparsity":[123],"objective,":[124],"enabling":[125],"context-specific":[126],"calibration":[127,187],"without":[128],"updating":[129],"any":[130],"backbone":[131],"parameters.":[132],"To":[133],"further":[134],"ground":[135],"visual":[139],"evidence,":[140],"we":[141],"incorporate":[142],"covariance-aware":[144],"Mahalanobis":[145],"refinement":[146],"stabilises":[148],"cross-modal":[149],"alignment.":[150],"Across":[151],"four":[152],"benchmarks,":[153],"MM-VAD":[154],"consistently":[155],"improves":[156],"prior":[158],"training-free":[159,199],"methods,":[160],"achieving":[161],"90.03%":[162],"AUC":[163],"XD-Violence":[165],"83.24%,":[167],"96.95%,":[168],"98.81%":[170],"UCF-Crime,":[172],"ShanghaiTech,":[173],"UCSD":[175],"Ped2,":[176],"respectively.":[177],"results":[179],"demonstrate":[180],"representation":[183],"provide":[188],"principled":[190],"effective":[192],"VAD.":[200]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-03-18T00:00:00"}