{"as_of":"2026-08-07T08:41:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:748ba49490b8b8263c2e81fec17f5225c5854390a63063e9bcd28c0eff6efaa4","coverage":[{"denominator":241,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":100,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-06T12:34:00.525559Z","state":"measured"},{"denominator":101,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":101,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-07T06:34:17.273281+00:00","state":"measured"},{"denominator":1,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":1,"source":"paper_references, paper_reference_links","source_observed_at":"2026-05-08T12:49:53.645987Z","state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"arxiv_reference","source_observed_at":"2026-05-11T19:01:19.377020Z","state":"measured"}],"external_citation_measurements":[],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"cited_work":{"arxiv_id":"2507.21649","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2507.21649","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"arXiv preprint arXiv:2507.21649 (2025)","venue":null,"work_id":"44ceaa4e-ad65-4eb1-9b00-ae8daced742d","year":2025},"citing_paper":{"arxiv_id":"2604.22884","last_updated":"2026-04-24T08:13:19Z","snapshot_observed_at":"2026-07-06T23:09:10.050398Z","submitted_at":"2026-04-24T08:13:19Z","title":"Can Multimodal Large Language Models Truly Understand Small Objects?","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-05-08T12:49:53.645987Z"},"links":{"cited_paper":"/paper/2507.21649","citing_paper":"/paper/2604.22884"},"observation_digest":"sha256:35e9f7d4d469b6ff8178bf88a8305273beca85033d4feae435a5adbc65c6a1f4","observation_id":"34432e23-5e8f-406e-a375-d4ec2927fcdc","resolution":{"observed_at":"2026-05-11T19:01:19.379185Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}}],"links":{"evidence":"/evidence","html":"/paper/2507.21649/citation-record","integrity":"/paper/2507.21649/integrity","json":"/paper/2507.21649/citation-record.json","paper":"/paper/2507.21649"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.159395Z","title":"Applications of outlier analysis,","venue":null,"work_id":null,"year":2012},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.159395Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:3a40ff8bf0ca62d6e757a5204c1cda7b87e8611d6223df63b74fb646f2f5470d","observation_id":"c45a6acb-ff9d-4c7b-9bd6-d6edd8cc887d","resolution":{"observed_at":"2026-08-06T12:34:00.159395Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2302.04549","last_updated":"2023-02-09T10:27:21Z","snapshot_observed_at":"2026-07-06T14:50:00.101378Z","submitted_at":"2023-02-09T10:27:21Z","title":"Weakly Supervised Anomaly Detection: A Survey","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2302.04549","snapshot_observed_at":"2026-08-06T12:34:00.162136Z","title":"Weakly supervised anomaly detection: A survey,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.162136Z"},"links":{"cited_paper":"/paper/2302.04549","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:9e5e326c6b3d8d503cc3a40ba4f1b709a0e88a00afca76ed03856d47e0d45d05","observation_id":"6f9a20fa-7c9a-490c-b928-338ca8bb7604","resolution":{"observed_at":"2026-08-06T12:34:00.162136Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.165334Z","title":"Deep learning for anomaly detection: A review,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.165334Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:8551fb49f27e56bc06c50360202c3db7e6baa3079b0e50ac87173e6be51fd9e5","observation_id":"4ba4aa76-f8f7-418b-9bb7-b2ff8fbb3f17","resolution":{"observed_at":"2026-08-06T12:34:00.165334Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.168148Z","title":"Dota: Unsupervised detection of traffic anomaly in driving videos,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.168148Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:df686c04d97c2b72aad2676bf82d0abd63d3ea45c1b8dd5c4fd071dcb04ee4fe","observation_id":"d3cd4329-4dc0-4c04-87e7-86368d8cf71e","resolution":{"observed_at":"2026-08-06T12:34:00.168148Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.170860Z","title":"Sparse reconstruction cost for abnormal event detection,","venue":null,"work_id":null,"year":2011},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.170860Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:42b5635f9ec6dcb79b0b51bae8f805526218cf4c650e80b593c35aa297ab71f5","observation_id":"d69239f0-7f34-4f99-8a90-da283eb404f5","resolution":{"observed_at":"2026-08-06T12:34:00.170860Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.173912Z","title":"Learning temporal regularity in video sequences,","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.173912Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:a6f4c9e2854ec434559ec31039f39d11c26c45b955233d29564a0f99fa91fd73","observation_id":"f9c0b77a-6b98-4db7-9823-dcde5747360f","resolution":{"observed_at":"2026-08-06T12:34:00.173912Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.176548Z","title":"Future frame prediction for anomaly detection–a new baseline,","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.176548Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:0b01419a5ec5eb3020bc37985ba127295ad8001fbd05834eced9945484a0140b","observation_id":"f62546e5-d22a-4ed6-8cf8-76ec8021a4bb","resolution":{"observed_at":"2026-08-06T12:34:00.176548Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.181866Z","title":"Uncovering what why and how: A compre- hensive benchmark for causation understanding of video anomaly,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.181866Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:0622f117417da1a833c2d42f83ec8a7e5673b74f6500c58ec08a861ca3ed2f82","observation_id":"ec3946a4-db25-4d6a-8404-e50b8f3f59f1","resolution":{"observed_at":"2026-08-06T12:34:00.181866Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.184451Z","title":"Hawk: Learning to understand open-world video anomalies,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.184451Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:b68751490316755b837de514697bdd41b37420222277f865e6255c11759824e1","observation_id":"a496e6cd-b3cb-4a29-b0d3-f4163cd7b07b","resolution":{"observed_at":"2026-08-06T12:34:00.184451Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.187120Z","title":"Vera: Explainable video anomaly detection via verbalized learning of vision-language models,","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.187120Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:a66ce5a2007bfc7dada9ee69078be6926bf555941251e22fc8627d775c18f7d3","observation_id":"202293b4-14f8-46e8-896e-cf398c53cf27","resolution":{"observed_at":"2026-08-06T12:34:00.187120Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.189781Z","title":"Holmes-vau: Towards long-term video anomaly understanding at any granularity,","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.189781Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:75f167db7320360a34bb3c6e8552ecf6b80080071c5511bf9b8b4c7827123030","observation_id":"1287bfae-b620-43d0-8392-b49828ce3438","resolution":{"observed_at":"2026-08-06T12:34:00.189781Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.192361Z","title":"Abnormal event detection at 150 fps in matlab,","venue":null,"work_id":null,"year":2013},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.192361Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:1bb9f13fb1b02774b8650158761cb541379a9d205d44214a4da922ddce158e5d","observation_id":"f058c997-81a4-4b30-af05-30ebf44ec2ea","resolution":{"observed_at":"2026-08-06T12:34:00.192361Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.195025Z","title":"Anomaly detection and localization in crowded scenes,","venue":null,"work_id":null,"year":2013},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.195025Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:d52f5a92273ef4b7ee9461965940d0a992bf044626e7135430e35d610acb9eca","observation_id":"4c95d730-5d89-4f3a-a5bc-2fcdd98aa500","resolution":{"observed_at":"2026-08-06T12:34:00.195025Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.197681Z","title":"Real-world anomaly detection in surveillance videos,","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.197681Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:ccf6208121cbbb681a6a40ad8648b3733cb6d8afbfd060a42fa10a04fbd62cc2","observation_id":"2f415b88-8f5d-4506-b33c-7dc72e2f1f25","resolution":{"observed_at":"2026-08-06T12:34:00.197681Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.205908Z","title":"A survey of single- scene video anomaly detection,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.205908Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:e22455a8e2cd5d581cc0239b14ddfb15f16378eae82f7598a047fb765f130820","observation_id":"89ab1ca4-8cbb-4114-b20d-30ae0d149732","resolution":{"observed_at":"2026-08-06T12:34:00.205908Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.208782Z","title":"A comprehensive review on deep learning-based methods for video anomaly detection,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.208782Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:577c55613ea8f05e633afc2aead16737a0aba84ea3a2ff259d1b1a87e00dbb62","observation_id":"d7e4ba7b-9ef2-468e-83d6-1dfba9dc4108","resolution":{"observed_at":"2026-08-06T12:34:00.208782Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.211386Z","title":"Anomaly analysis in images and videos: A comprehensive review,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.211386Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:a4283179a722ff140217e3533376f1dffc8e42a2c1063675ffecaf807e1225fd","observation_id":"fe7a0fe2-ab32-4515-8ddf-e46c0926ff2a","resolution":{"observed_at":"2026-08-06T12:34:00.211386Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.213975Z","title":"Generalized video anomaly event detection: Systematic taxonomy and comparison of deep models,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.213975Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:94a9598b51e2972c7888eb8ea35375254ac0815f0d845ad1b7c35b7cef4e92f8","observation_id":"6d5084c6-9a25-4687-bb6b-9c2cd9f2e8a6","resolution":{"observed_at":"2026-08-06T12:34:00.213975Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2409.05383","last_updated":"2024-09-09T07:31:16Z","snapshot_observed_at":"2026-07-06T19:12:23.792177Z","submitted_at":"2024-09-09T07:31:16Z","title":"Deep Learning for Video Anomaly Detection: A Review","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2409.05383","snapshot_observed_at":"2026-08-06T12:34:00.216618Z","title":"Deep learning for video anomaly detection: A review,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.216618Z"},"links":{"cited_paper":"/paper/2409.05383","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:3b20f520ccf50d87e82cc24cdd8768eefc90c4af39bbeba851a164a28f5b06d8","observation_id":"c7ee42d3-2f38-4557-b071-e817ba8dab60","resolution":{"observed_at":"2026-08-06T12:34:00.216618Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2405.19387","last_updated":"2024-07-01T02:31:53Z","snapshot_observed_at":"2026-07-06T18:22:14.521977Z","submitted_at":"2024-05-29T17:56:31Z","title":"Video Anomaly Detection in 10 Years: A Survey and Outlook","version":2},"cited_work":{"arxiv_id":"2405.19387","doi":null,"metadata_source":"pith","pith_arxiv_id":"2405.19387","snapshot_observed_at":"2026-08-06T12:34:01.530908Z","title":"Video Anomaly Detection in 10 Years: A Survey and Outlook","venue":"cs.CV","work_id":"bac53c31-956a-4877-a8d8-9d66d7ac7504","year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.219529Z"},"links":{"cited_paper":"/paper/2405.19387","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:e05ccae388059edac8616cbb73cc78da2c133e98eaf1edd9bf622bde6181304d","observation_id":"41e0b66f-7409-4ccb-a4dc-bafcea50f6b4","resolution":{"observed_at":"2026-08-06T12:34:01.534149Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2412.18298","last_updated":"2024-12-24T09:05:37Z","snapshot_observed_at":"2026-07-06T20:12:40.901520Z","submitted_at":"2024-12-24T09:05:37Z","title":"Quo Vadis, Anomaly Detection? LLMs and VLMs in the Spotlight","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2412.18298","snapshot_observed_at":"2026-08-06T12:34:00.222808Z","title":"Quo vadis, anomaly detection? llms and vlms in the spotlight,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.222808Z"},"links":{"cited_paper":"/paper/2412.18298","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:be27c301bb46e9b8ec2db8f3a56fd5086913ce089e7c1fbb076d0f6892f6ca8a","observation_id":"c31c3a18-cde0-4b82-8042-fb614ab1f685","resolution":{"observed_at":"2026-08-06T12:34:00.222808Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.225961Z","title":"Networking systems for video anomaly detection: A tutorial and survey,","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.225961Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:4649629faa95b0bf80deb5042e8655b3d4e989472726d4ce19e8632627d4ed85","observation_id":"7cdf79dd-fab1-4c23-8153-53a2e96c637d","resolution":{"observed_at":"2026-08-06T12:34:00.225961Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2110.09742","last_updated":"2021-10-24T06:43:21Z","snapshot_observed_at":"2026-07-06T11:59:22.353830Z","submitted_at":"2021-10-19T05:22:38Z","title":"Learning Not to Reconstruct Anomalies","version":2},"cited_work":{"arxiv_id":"2110.09742","doi":null,"metadata_source":"pith","pith_arxiv_id":"2110.09742","snapshot_observed_at":"2026-08-06T12:34:01.512012Z","title":"Learning Not to Reconstruct Anomalies","venue":"cs.CV","work_id":"5c346aa8-afaf-4239-af50-725c7efc18a7","year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.228618Z"},"links":{"cited_paper":"/paper/2110.09742","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:c7281db5083dd0c72ee12fe356b43d098612a31bb57391ddc9c816fc6c320770","observation_id":"b4d1dd38-1086-4932-9d26-db6a5a3c2292","resolution":{"observed_at":"2026-08-06T12:34:01.515218Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.231599Z","title":"Making reconstruction-based method great again for video anomaly detection,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.231599Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:fd05b711f18e04f280b664ced8834509b5e1f814607b5bcab39d543b7ea99554","observation_id":"ee4a3715-1386-4207-892b-62b26cd2f159","resolution":{"observed_at":"2026-08-06T12:34:00.231599Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2211.10052","last_updated":"2022-11-28T13:02:01Z","snapshot_observed_at":"2026-07-06T14:20:06.936367Z","submitted_at":"2022-11-18T06:41:02Z","title":"Pedestrian Spatio-Temporal Information Fusion For Video Anomaly Detection","version":2},"cited_work":{"arxiv_id":"2211.10052","doi":null,"metadata_source":"pith","pith_arxiv_id":"2211.10052","snapshot_observed_at":"2026-08-06T12:34:01.500371Z","title":"Pedestrian Spatio-Temporal Information Fusion For Video Anomaly Detection","venue":"cs.CV","work_id":"1307447a-bd9c-414d-9ac0-7f7961c110b6","year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.237068Z"},"links":{"cited_paper":"/paper/2211.10052","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:2b907056e817644e46eac0728bd1a89adb654b408bda8aa4009f5d4ff5537286","observation_id":"7472c54e-4265-408c-bfdf-ba356ed92c24","resolution":{"observed_at":"2026-08-06T12:34:01.503824Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1907.10211","last_updated":"2019-07-24T02:36:28Z","snapshot_observed_at":"2026-07-06T08:09:45.736165Z","submitted_at":"2019-07-24T02:36:28Z","title":"Motion-Aware Feature for Improved Video Anomaly Detection","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1907.10211","snapshot_observed_at":"2026-08-06T12:34:00.240035Z","title":"Motion-aware feature for improved video anomaly detection,","venue":null,"work_id":null,"year":1907},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.240035Z"},"links":{"cited_paper":"/paper/1907.10211","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:bd0bfd09dde171be3f41ae3e766cf910c3b4b2365ef9db6b26ab8f2825ab9e86","observation_id":"c5f1f3ff-7597-44a3-be8a-9997abd02fcb","resolution":{"observed_at":"2026-08-06T12:34:00.240035Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.243068Z","title":"Temporal convolutional network with complementary inner bag loss for weakly supervised anomaly detec- tion,","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.243068Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:a18aa73148d464020938d42cb92876e69d6b4f014081576f0e8d7e3a2b6cd720","observation_id":"b1a93c27-615c-4cfb-a816-055f75b75489","resolution":{"observed_at":"2026-08-06T12:34:00.243068Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.245722Z","title":"Collaborative normality learning framework for weakly supervised video anomaly detection,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.245722Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:41d70452788e1eeb52c2d2142374f57c5fb5ecb3a2154a9baa7974c2da4ae962","observation_id":"61ac0e72-4f8d-4133-8782-1ce968a5b444","resolution":{"observed_at":"2026-08-06T12:34:00.245722Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.248495Z","title":"Vadclip: Adapting vision-language models for weakly supervised video anomaly detection,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.248495Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:b2da5e16025a6cdce9f73eb2a0b46802dadc74bc4c6e5cdf1c7573284cdee459","observation_id":"6b2cc5e3-23fe-42aa-bfaf-24798fa54a01","resolution":{"observed_at":"2026-08-06T12:34:00.248495Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.251585Z","title":"Harnessing large language models for training-free video anomaly detection,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.251585Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:34d0e3b49d322ab89d5bdf94cfad3c15de8084e838d4311b92ed2c0e5bab50ad","observation_id":"aa4b18fe-5867-4710-97ae-2a6f99f5b0f0","resolution":{"observed_at":"2026-08-06T12:34:00.251585Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.254465Z","title":"Follow the rules: reasoning for video anomaly detection with large language models,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.254465Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:4e7a7f59432d34fd592a07314bbddde684c0cf8c758619a2beec650a98119ebf","observation_id":"963230db-5875-48b7-b936-9ed360839722","resolution":{"observed_at":"2026-08-06T12:34:00.254465Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2302.13971","last_updated":"2023-02-27T17:11:15Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-02-27T17:11:15Z","title":"LLaMA: Open and Efficient Foundation Language Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2302.13971","snapshot_observed_at":"2026-08-06T12:34:00.257306Z","title":"Llama: Open and efficient foundation language models,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.257306Z"},"links":{"cited_paper":"/paper/2302.13971","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:3c72de5cf306d56f18c6f9563de584d6f8cd70b6bd05c29202c6b419e8e54c85","observation_id":"c021adb2-078c-4d05-ae07-dd61fa5ca5f0","resolution":{"observed_at":"2026-08-06T12:34:00.257306Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.260231Z","title":"Internvl: Scaling up vision foundation models and aligning for generic visual-linguistic tasks,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.260231Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:e3ff25542787019ab186804fbc9b188f59f884e85c0749fdd8ac791315bb1eed","observation_id":"abff9647-fc4a-4c8b-9c3f-2196e4c379f1","resolution":{"observed_at":"2026-08-06T12:34:00.260231Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.262755Z","title":"Suvad: Semantic understanding based video anomaly detection using mllm,","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.262755Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:82ccc425bccdcc4522eb0e0a776739fc7448ad9744ff290fadb4868d5b5e0d69","observation_id":"1d824e62-c36d-4cd1-bd78-4a99deb67137","resolution":{"observed_at":"2026-08-06T12:34:00.262755Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.265630Z","title":"Robust real-time unusual event detection using multiple fixed-location monitors,","venue":null,"work_id":null,"year":2008},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.265630Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:f957ff7ec98e4162b093da30ed33242fc44a4b1fb27d5f3095e196158aa90bcc","observation_id":"560d6c1b-c95d-4efe-8766-a213b742f81a","resolution":{"observed_at":"2026-08-06T12:34:00.265630Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.268383Z","title":"Abnormal crowd behavior detection using social force model,","venue":null,"work_id":null,"year":2009},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.268383Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:b5ebbaf6550ad84e93d4072d1852afe25cb1fa0782d3fa9493188b7c63a86e6c","observation_id":"4e646852-8ac3-40bf-991d-ef135248e99d","resolution":{"observed_at":"2026-08-06T12:34:00.268383Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.271234Z","title":"Street scene: A new dataset and evaluation protocol for video anomaly detection,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.271234Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:fd76acfb288fa3951d69eccf0d912aed7804e6f1f900dc9a40f0f53f34eeff3b","observation_id":"40bab5f6-139d-44bc-bab8-01679b86db6c","resolution":{"observed_at":"2026-08-06T12:34:00.271234Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.273932Z","title":"A new comprehensive bench- mark for semi-supervised video anomaly detection and anticipation,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.273932Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:09be39644257171759bbef08d3cf4a70f7741e1c2553ac77a7676907d02f06cf","observation_id":"45268180-1c93-496a-bf6a-23eafb537ed9","resolution":{"observed_at":"2026-08-06T12:34:00.273932Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.276814Z","title":"A revisit of sparse coding based anomaly detection in stacked rnn framework,","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.276814Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:cbea9d979077ce3b4c1a697ac382ef233ae1caa66a626119e7a591f3fe31b12a","observation_id":"74ebf29a-8cdc-448a-8ded-7cb665231903","resolution":{"observed_at":"2026-08-06T12:34:00.276814Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.279688Z","title":"Adnet: Temporal anomaly detection in surveillance videos,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.279688Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:6ca0af9dcd1918cfca31a8ff2ad918f4232f237557523e38996d6deaa1b1f8e7","observation_id":"2c6cc3c7-2c80-41a3-8830-8745cfc41631","resolution":{"observed_at":"2026-08-06T12:34:00.279688Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.282403Z","title":"Tad: A large-scale benchmark for traffic accidents detection from video surveillance,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.282403Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:27a070534ca225b5b0e089d6f0a28dd602d019c70a4d056c1c9c1dbfb07765b4","observation_id":"2f38967c-40bb-4944-a506-b43718e4371e","resolution":{"observed_at":"2026-08-06T12:34:00.282403Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.285181Z","title":"People detection and pose classification inside a moving train using computer vision,","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.285181Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:7c7ea5e06e753efaf15c5b54bbfb59239707e42ed9096cfdc4181bb672827dc1","observation_id":"b93ae489-6b91-44fb-a6ec-84c6d32370ad","resolution":{"observed_at":"2026-08-06T12:34:00.285181Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.288704Z","title":"Camnuvem: A robbery dataset for video anomaly detection,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.288704Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:62b3f10f264e8a4721d8fff2eac563f53b1ef78f7e7ec8bb3076465614e95813","observation_id":"c68bb1d7-1c57-472b-8b05-091012981a09","resolution":{"observed_at":"2026-08-06T12:34:00.288704Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2502.09325","last_updated":"2025-02-13T13:38:17Z","snapshot_observed_at":"2026-08-04T01:21:26.867219Z","submitted_at":"2025-02-13T13:38:17Z","title":"A Benchmark for Crime Surveillance Video Analysis with Large Models","version":1},"cited_work":{"arxiv_id":"2502.09325","doi":null,"metadata_source":"pith","pith_arxiv_id":"2502.09325","snapshot_observed_at":"2026-08-06T12:34:01.473793Z","title":"A Benchmark for Crime Surveillance Video Analysis with Large Models","venue":"cs.CV","work_id":"6849f534-a3ee-4035-9b5b-fdf5cd64821b","year":2025},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.291412Z"},"links":{"cited_paper":"/paper/2502.09325","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:8037a294352ae6bb6e7fe4d245d1a56e7afbfdd54b9d9322574d2a7da9d6a317","observation_id":"5c9d601a-255d-4600-98bf-4cd34de9293f","resolution":{"observed_at":"2026-08-06T12:34:01.477068Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2004.03044","last_updated":"2020-04-06T23:58:59Z","snapshot_observed_at":"2026-07-06T09:10:26.590798Z","submitted_at":"2020-04-06T23:58:59Z","title":"When, Where, and What? A New Dataset for Anomaly Detection in Driving Videos","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2004.03044","snapshot_observed_at":"2026-08-06T12:34:00.294526Z","title":"When, where, and what? a new dataset for anomaly detection in driving videos,","venue":null,"work_id":null,"year":2004},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":50,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.294526Z"},"links":{"cited_paper":"/paper/2004.03044","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:c8b5634184d0af52d49756d69b4b0222ac91d101c5b302470a87329ec325ff7e","observation_id":"3aa377e1-dcd3-4d29-baa8-615c4ea38524","resolution":{"observed_at":"2026-08-06T12:34:00.294526Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.297527Z","title":"Ubnormal: New benchmark for supervised open-set video anomaly detection,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":51,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.297527Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:e0c6fb0b255f4d8115cb5da3ce230c58282f4333ec558e20f4776c120b3b92a7","observation_id":"d568bbd0-719e-4000-ae53-92347cb46694","resolution":{"observed_at":"2026-08-06T12:34:00.297527Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2412.07183","last_updated":"2024-12-10T04:41:44Z","snapshot_observed_at":"2026-07-06T20:04:25.198952Z","submitted_at":"2024-12-10T04:41:44Z","title":"Exploring What Why and How: A Multifaceted Benchmark for Causation Understanding of Video Anomaly","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2412.07183","snapshot_observed_at":"2026-08-06T12:34:00.300356Z","title":"Exploring what why and how: A multifaceted benchmark for causation understanding of video anomaly,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":52,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.300356Z"},"links":{"cited_paper":"/paper/2412.07183","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:1410ac991142500f2010e9b4f88ef460c701f511dc422a6050a7ce2e071e73e8","observation_id":"430a411e-5753-4982-8be5-58ca4ecc59f0","resolution":{"observed_at":"2026-08-06T12:34:00.300356Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2406.10326","last_updated":"2025-03-24T20:26:56Z","snapshot_observed_at":"2026-07-06T18:31:13.039726Z","submitted_at":"2024-06-14T17:59:01Z","title":"VANE-Bench: Video Anomaly Evaluation Benchmark for Conversational LMMs","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.10326","snapshot_observed_at":"2026-08-06T12:34:00.303325Z","title":"Vane- bench: Video anomaly evaluation benchmark for conversational lmms,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":53,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.303325Z"},"links":{"cited_paper":"/paper/2406.10326","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:6a53941dc374aa1705ee5e59999e624a3adba3069968f307216898ea4a3d8fad","observation_id":"452d992d-51e3-4788-b382-d3920acd75fe","resolution":{"observed_at":"2026-08-06T12:34:00.303325Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.309363Z","title":"Towards surveillance video-and-language understanding: New dataset baselines and challenges,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":55,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.309363Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:d245f3d2cd28ad74600b5610200530d46caeb335455e3610a3eb762b47ada8fa","observation_id":"4017d836-9e21-4d66-a2e5-911ab0c451ec","resolution":{"observed_at":"2026-08-06T12:34:00.309363Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.311978Z","title":"Two-person interaction detection using body-pose features and mul- tiple instance learning,","venue":null,"work_id":null,"year":2012},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":56,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.311978Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:e4c373219058afa2108eb5ffcfbc71c595556328d165084a4b558c4b19a34194","observation_id":"6353174f-7983-476d-8292-d77b4bdd8693","resolution":{"observed_at":"2026-08-06T12:34:00.311978Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.314566Z","title":"Bleu: a method for automatic evaluation of machine translation,","venue":null,"work_id":null,"year":2002},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":57,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.314566Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:971a1d387c2faa12914039a4ca6e4d3abdf494869a003bb60fad9c7cc8c92243","observation_id":"10fec61d-32c9-4146-88bc-27feb953ae4f","resolution":{"observed_at":"2026-08-06T12:34:00.314566Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.317469Z","title":"Rouge: A package for automatic evaluation of summaries,","venue":null,"work_id":null,"year":2004},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":58,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.317469Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:184c60949e95d1c53edc1fb5648128e9218222146cc0d2e51e4e90045347def3","observation_id":"c4cc34a3-7458-4241-a3a5-2588767360b5","resolution":{"observed_at":"2026-08-06T12:34:00.317469Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.320150Z","title":"Meteor: An automatic metric for mt evalua- tion with improved correlation with human judgments,","venue":null,"work_id":null,"year":2005},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":59,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.320150Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:fa8d223d0a0e459cef06bf1cf808cfc5fc81a0b0036c729f071fd759deb7b1f4","observation_id":"757a96f4-955d-4175-b77d-031ffaf434e5","resolution":{"observed_at":"2026-08-06T12:34:00.320150Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.323155Z","title":"Self- supervised sparse representation for video anomaly detection,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":60,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.323155Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:9f3d7d866d7597dffd64f123e284540e5da2a936773151eb0953623ab42419e4","observation_id":"dad5e77a-166f-474f-b464-4c6d0920b266","resolution":{"observed_at":"2026-08-06T12:34:00.323155Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.325893Z","title":"Self-training multi-sequence learning with transformer for weakly supervised video anomaly detection,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":61,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.325893Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:44d48ccec14d648abd7da47d4b48e5ef12d5356d5fa5df3e052c0c4cb206886e","observation_id":"92a0fb1c-1223-49d4-b626-09004343ec90","resolution":{"observed_at":"2026-08-06T12:34:00.325893Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.329416Z","title":"Exploiting completeness and uncertainty of pseudo labels for weakly supervised video anomaly detection,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":62,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.329416Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:3d29af1089cae74068f8d0d9bd03a449dabb88a0142404f626af26aaca8c844c","observation_id":"98ec9b9b-f4ce-4295-8783-44ec361e052f","resolution":{"observed_at":"2026-08-06T12:34:00.329416Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.332235Z","title":"Learning causal temporal relation and feature discrimination for anomaly detection,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":63,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.332235Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:de8066d5defdc05b6cd0cde0695d7850bb7c87a0db69528491ee9986c80dfc30","observation_id":"1946af3a-5fd8-4f30-8cb7-f7a0dba68334","resolution":{"observed_at":"2026-08-06T12:34:00.332235Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.335189Z","title":"Weakly-supervised video anomaly detection with robust temporal feature magnitude learning,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":64,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.335189Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:81f81266acf7d4350ced7d169b3f6aef1e723b3ed6744761d29f5f505156c802","observation_id":"49e64bcf-3e1d-4096-96b3-1a3f6d47d511","resolution":{"observed_at":"2026-08-06T12:34:00.335189Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.337878Z","title":"Look around for anomalies: Weakly-supervised anomaly detection via context- motion relational learning,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":65,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.337878Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:23bb8e771956bd2741a2e43559566f784d93f8c7e2010424cf00f0ec240c1e12","observation_id":"a605e787-1434-4343-a99f-1bc192809475","resolution":{"observed_at":"2026-08-06T12:34:00.337878Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.340671Z","title":"Mgfn: Magnitude-contrastive glance-and-focus network for weakly- supervised video anomaly detection,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":66,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.340671Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:662c18cf6687811a02a269a9af801e93277a0f755113e304a019426cbf40e664","observation_id":"92a2959f-620b-40ee-ac74-18f120f1a7ad","resolution":{"observed_at":"2026-08-06T12:34:00.340671Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.343595Z","title":"Quo vadis, action recognition? a new model and the kinetics dataset,","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":67,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.343595Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:6c7bf63a129ba250a63b030cac3048fe063851c33f067d7224b25868f02249c9","observation_id":"c63cc95d-7ed8-4e69-8f2b-1ac7940ac97b","resolution":{"observed_at":"2026-08-06T12:34:00.343595Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.346330Z","title":"Video swin transformer,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":68,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.346330Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:485eb55805ca000294c4580a0423ff1d348dcb0738675efc08a5811813e6e138","observation_id":"2272b5b4-4195-4d4a-8811-21b7d737ce25","resolution":{"observed_at":"2026-08-06T12:34:00.346330Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.349264Z","title":"Text prompt with normality guidance for weakly supervised video anomaly detection,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":69,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.349264Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:b1d5901202bb5ea6c457406c84256978bcc5738341f87dc4670c2f8cd425aef6","observation_id":"cb18144a-76b6-4654-a077-055158463cbe","resolution":{"observed_at":"2026-08-06T12:34:00.349264Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.351978Z","title":"Unbiased multi- ple instance learning for weakly supervised video anomaly detection,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":70,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.351978Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:c15674ef393f3e068b706fd8afbb14e1130153831b5f23fe516824a1f952b96b","observation_id":"381f0e1a-75d4-4166-8321-a9d7885d15dd","resolution":{"observed_at":"2026-08-06T12:34:00.351978Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.355001Z","title":"Weakly supervised video anomaly detection and localization with spatio-temporal prompts,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":71,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.355001Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:1fd9320875dcfa7157655c1848fa3fa424b2d77cfa0be8d217602bdc06744083","observation_id":"ff1ded7e-8932-4b85-8473-cd605b40f2a8","resolution":{"observed_at":"2026-08-06T12:34:00.355001Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2409.14109","last_updated":"2024-09-26T01:38:52Z","snapshot_observed_at":"2026-07-06T19:19:11.152731Z","submitted_at":"2024-09-21T11:48:54Z","title":"Vision-Language Models Assisted Unsupervised Video Anomaly Detection","version":2},"cited_work":{"arxiv_id":"2409.14109","doi":null,"metadata_source":"pith","pith_arxiv_id":"2409.14109","snapshot_observed_at":"2026-08-06T12:34:01.374311Z","title":"Vision-Language Models Assisted Unsupervised Video Anomaly Detection","venue":"cs.CV","work_id":"2c6d29a2-0688-469e-9768-4c7c64c8f3ca","year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":72,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.357795Z"},"links":{"cited_paper":"/paper/2409.14109","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:f42ea2d7b5726449a2972cb5aaa518025f44094b7af985ad5c5d48185ab3c510","observation_id":"1a8ccc85-663e-4be8-a1c6-0d471cfab4e7","resolution":{"observed_at":"2026-08-06T12:34:01.378212Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2401.05702","last_updated":"2024-01-11T07:09:44Z","snapshot_observed_at":"2026-07-06T17:14:07.183954Z","submitted_at":"2024-01-11T07:09:44Z","title":"Video Anomaly Detection and Explanation via Large Language Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2401.05702","snapshot_observed_at":"2026-08-06T12:34:00.360891Z","title":"Video anomaly detection and explanation via large language models,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":73,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.360891Z"},"links":{"cited_paper":"/paper/2401.05702","citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:16de63401a236784e1b85e35c0261587e89e3e6bfea99f8d9f9ba90fbc1015cb","observation_id":"73ca25d3-75a5-496e-adec-9ab50552daa1","resolution":{"observed_at":"2026-08-06T12:34:00.360891Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.364739Z","title":"Mist: Multi-modal iterative spatial-temporal transformer for long-form video question answering,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":74,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.364739Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:4f987c2986c09a64ac03b864dc4e2c9d3099bb65534dbe9d68ff3cc2a2a44787","observation_id":"ea21770b-bf58-4bb9-946b-d5b07d57dc38","resolution":{"observed_at":"2026-08-06T12:34:00.364739Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.367524Z","title":"Object- centric auto-encoders and dummy anomalies for abnormal event detec- tion in video,","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":75,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.367524Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:c4738c8eed235bb21d979ab525bc1c9a2e8e353120da2019a8c51e02cc13df9e","observation_id":"4e87cc5c-1929-4ec5-88e8-950c07881f3f","resolution":{"observed_at":"2026-08-06T12:34:00.367524Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.373016Z","title":"A background-agnostic framework with adversarial training for abnor- mal event detection in video,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":77,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.373016Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:2f1d5e18d59a0ed3acc6d689e51a9859c837b30716b4a2d8d15b2e3fa3816001","observation_id":"d8f4b07e-5e49-40d1-9b3f-34fea95d0268","resolution":{"observed_at":"2026-08-06T12:34:00.373016Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.375931Z","title":"Anomaly detection in video se- quence with appearance-motion correspondence,","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":78,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.375931Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:7771a402ab77f1d37d18c1c3b6ca6516accab33a2c0e03cc024aacf5d2e4c438","observation_id":"a96cba92-696f-4b8d-ad0b-43257906707c","resolution":{"observed_at":"2026-08-06T12:34:00.375931Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.378754Z","title":"Generative neural networks for anomaly detection in crowded scenes,","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":79,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.378754Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:fd71b4bb3befed88a96b95e71ed376be222da28135c5f81a06642e4433f163c9","observation_id":"f2670b79-0e67-4ac3-ad11-51aa864e5d3d","resolution":{"observed_at":"2026-08-06T12:34:00.378754Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.381641Z","title":"Regularity learning via explicit distribution modeling for skeletal video anomaly detection,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":80,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.381641Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:bfce8759c6b6d2e245d296ded05425d0009e2daa1527980c95f6acb729951cdf","observation_id":"817f4573-5a0e-48a7-ab39-93ad27169470","resolution":{"observed_at":"2026-08-06T12:34:00.381641Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.384359Z","title":"Clustering driven deep autoencoder for video anomaly detection,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":81,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.384359Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:536b1fcb74ae9de307370a24a82419069d3dc28c148e1165ed4ccc2fde6942a9","observation_id":"6f6b62eb-187f-4732-888a-8c30c3955322","resolution":{"observed_at":"2026-08-06T12:34:00.384359Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.387649Z","title":"Anomaly detection with bidirectional consistency in videos,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":82,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.387649Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:063ca251c0eaf5da5fec12bd66b0fb1b23b5d964a4e2be956818e8f721b8a9b3","observation_id":"9cc4954a-7e7d-45cd-be05-7f7ff2420aed","resolution":{"observed_at":"2026-08-06T12:34:00.387649Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.390377Z","title":"Self-supervision-augmented deep autoencoder for unsupervised visual anomaly detection,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":83,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.390377Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:a2a0c45c28854b810622a8f877fe0f89e5249ef0f9b51ae1901139a7b719d70a","observation_id":"dd10a135-ae1c-4b16-b585-d0e4b17cbea2","resolution":{"observed_at":"2026-08-06T12:34:00.390377Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.393121Z","title":"Remembering history with convolutional lstm for anomaly detection,","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":84,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.393121Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:d7b6f99930f83e4ccb2781ecfaef2e3d7e4699dc0b2330f184323b209f9e2441","observation_id":"50d414a7-45b0-4dd7-94f0-31af56f86f7c","resolution":{"observed_at":"2026-08-06T12:34:00.393121Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.395919Z","title":"A hybrid video anomaly detection framework via memory-augmented flow reconstruction and flow-guided frame prediction,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":85,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.395919Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:5d3d0ffe94b563d9aea9d06c03d20edc543dd0d1e60fcbf0c988846c123eccee","observation_id":"76cd1efe-855d-4d00-8fc3-d1781ffb22d2","resolution":{"observed_at":"2026-08-06T12:34:00.395919Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.398506Z","title":"Appearance-motion memory consistency network for video anomaly detection,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":86,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.398506Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:ed3cd0035e133ac6257f9212d0bcc67cc512bc93482800935da00f11a7333a45","observation_id":"44fd05e1-5ccb-407e-8b76-037df76e59da","resolution":{"observed_at":"2026-08-06T12:34:00.398506Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.401415Z","title":"Hierarchical graph embedded pose regularity learning via spatio- temporal transformer for abnormal behavior detection,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":87,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.401415Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:7f63b64d6d9ef35a7a57cd40846547197f5b9fa1476bbff91ccc06968aaba2cb","observation_id":"780d526d-16b2-4c3c-a598-f1b4e2f95e86","resolution":{"observed_at":"2026-08-06T12:34:00.401415Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.404428Z","title":"Attention- driven loss for anomaly detection in video surveillance,","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":88,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.404428Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:92813d552431e7479fa608dd3e154df3b57c796335a403ccf84d4f2dba1f78cd","observation_id":"6870c4a5-d9ea-4c79-8c25-c990d921f832","resolution":{"observed_at":"2026-08-06T12:34:00.404428Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.407326Z","title":"Normality learning in multispace for video anomaly detection,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":89,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.407326Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:42193841f6787ff5523f09bbe3f105d6f75a3b4fe88c4b770cc479b58a7c7fae","observation_id":"5bfffa8e-ac57-4b79-9b84-fbbfbf491392","resolution":{"observed_at":"2026-08-06T12:34:00.407326Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.410174Z","title":"Robust unsupervised video anomaly detection by multipath frame prediction,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":90,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.410174Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:a3a50634e73929d9e655a81dc656d3284865cbfd451d942839c8c1100877869d","observation_id":"7da9a2c6-ad91-4cf5-b761-70d3f4dcf126","resolution":{"observed_at":"2026-08-06T12:34:00.410174Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.413779Z","title":"Abnormal event detection and localization via adversarial event prediction,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":91,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.413779Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:1697ba865bfca57e544f2b9dcec3f4a07ddd767807f750e9e561a0c3e59f2973","observation_id":"b0c8f3be-23f4-4f48-b828-de3ffa4112c9","resolution":{"observed_at":"2026-08-06T12:34:00.413779Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.416531Z","title":"Object-guided and motion-refined atten- tion network for video anomaly detection,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":92,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.416531Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:6a1f6d531dc183995c52f3bee36b0e07020a301212cdd7b4133aae31da13efc0","observation_id":"a849242d-b4aa-4cb3-91d0-da7f0f0e1a17","resolution":{"observed_at":"2026-08-06T12:34:00.416531Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.419246Z","title":"Spatial- temporal graph convolutional network boosted flow-frame prediction for video anomaly detection,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":93,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.419246Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:3e3b655b66094f6762481200d2920e7cd63a75632e0b45df23033fd4ea00bd7a","observation_id":"f1f01806-f4cc-49e9-9031-de373a7f3181","resolution":{"observed_at":"2026-08-06T12:34:00.419246Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.421819Z","title":"Amp-net: Appearance-motion prototype network assisted automatic video anomaly detection system,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":94,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.421819Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:36a5afc05c8f1614184a456b4885cdeae0edb9fc34e5355e4c850233a36fcad8","observation_id":"e2b121fb-a50e-403f-ad2a-478e4009d042","resolution":{"observed_at":"2026-08-06T12:34:00.421819Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.424533Z","title":"Cloze test helps: Effective video anomaly detection via learning to complete video events,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":95,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.424533Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:710f2eff8ebde88e2f1820c3b437f49a0f930169134968a1d0807633e1cfc4f7","observation_id":"b47c9610-b361-4b17-a49e-9d045cdc64d1","resolution":{"observed_at":"2026-08-06T12:34:00.424533Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.427255Z","title":"Video event restoration based on keyframes for video anomaly detection,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":96,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.427255Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:15cfe3ae66fc67ffe83fabe6e9b400473c3809c9078d16d5bb953fd1aaa1ca3e","observation_id":"03d9a9b0-adae-4112-98da-d3bd15629c9b","resolution":{"observed_at":"2026-08-06T12:34:00.427255Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.429947Z","title":"Video anomaly detection via visual cloze tests,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":97,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.429947Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:2df5c2323db7295250803bdd74b9d0e3ba8841ab2dbc11fd18a04819669e5b1e","observation_id":"6a146e35-dbd8-44b4-9a1e-abde452310f8","resolution":{"observed_at":"2026-08-06T12:34:00.429947Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.433020Z","title":"Video anomaly detection by solving decoupled spatio-temporal jigsaw puzzles,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":98,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.433020Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:1586608fa9a7be5e379b3e4cdac8528594fbb12b8fc128f896a79a5eed1aba6f","observation_id":"45920885-f140-438c-bec9-a6ba72605b4d","resolution":{"observed_at":"2026-08-06T12:34:00.433020Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.435696Z","title":"Video anomaly detection via sequentially learning multiple pretext tasks,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":99,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.435696Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:d62673166cb89f90bc15c2ccef84118c2105784ac02533ba0dda659bdd282f2f","observation_id":"bc68fea4-3004-4419-bdc8-3e1be45b0074","resolution":{"observed_at":"2026-08-06T12:34:00.435696Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.438422Z","title":"Ssmtl++: Revisiting self-supervised multi-task learning for video anomaly detection,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":100,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.438422Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:99dc2e851127083f5d34dced83fac470db2d0484cff562c9376280e090906c51","observation_id":"9df77423-1fde-4440-8cc4-f9bbbcfce8f2","resolution":{"observed_at":"2026-08-06T12:34:00.438422Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.441191Z","title":"Abnormal event detection using deep contrastive learning for intelligent video surveillance system,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":101,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.441191Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:1ad73046b2ec2d8df47bc946eade8d2e75047fb6daaca34c870b5dc43cc25ffd","observation_id":"846e6845-bf7c-42a6-ad41-17c4e07c0cb5","resolution":{"observed_at":"2026-08-06T12:34:00.441191Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.443893Z","title":"Cluster attention contrast for video anomaly detection,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":102,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.443893Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:6965ff04a333da4209e4101dcfdac040ab3e1d3b8ea253bd9bc2d14ba7c7f336","observation_id":"c7ce6d6a-ceb6-4634-99f5-9369142f5edc","resolution":{"observed_at":"2026-08-06T12:34:00.443893Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.446615Z","title":"Learnable locality-sensitive hashing for video anomaly detection,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":103,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.446615Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:5b606844f5050e0017df5d15d3a98cd449f145ad476cd598742c4b9aa79979f0","observation_id":"eb0327ed-6c56-4efa-b3fe-987b35282ae4","resolution":{"observed_at":"2026-08-06T12:34:00.446615Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.449358Z","title":"Multimodal motion conditioned diffusion model for skeleton-based video anomaly detection,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":104,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.449358Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:c9d88c4d7d6abd2789314944cef504ba59f6251085095384b621703a11caa8a6","observation_id":"c6b67398-571b-4f45-8aa3-381cffc78031","resolution":{"observed_at":"2026-08-06T12:34:00.449358Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.452093Z","title":"Adversarial 3d convolutional auto- encoder for abnormal event detection in videos,","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":105,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.452093Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:dc17b27f57834df8ed6b283be30555cfa3ab5a5bbdb82a3be3574cdfc6ded423","observation_id":"66a08337-be92-4fa9-887b-9cc1f6389b62","resolution":{"observed_at":"2026-08-06T12:34:00.452093Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.522477Z","title":"Nm-gan: Noise- modulated generative adversarial network for video anomaly detec- tion,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":106,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.522477Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:505626708bccba5f60a69dd84601c3e3ee041df6a7b40111fa8e549154b320ec","observation_id":"3f1c0a88-efad-41e7-a9e7-433ba584b6be","resolution":{"observed_at":"2026-08-06T12:34:00.522477Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T12:34:00.525559Z","title":"Spatio- temporal autoencoder for video anomaly detection,","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM","version":1},"reference_index":107,"source":"pdf_text","source_observed_at":"2026-08-06T12:34:00.525559Z"},"links":{"citing_paper":"/paper/2507.21649"},"observation_digest":"sha256:a2971fad136dd26b79b9a0f026d8076520ab8a44429e5fc8639a5e2a06de1cd1","observation_id":"d9e904ff-55f5-47ab-8276-69218ef69135","resolution":{"observed_at":"2026-08-06T12:34:00.525559Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2507.21649","last_updated":"2025-07-29T10:07:24Z","latest_version":1,"primary_category":"cs.CV","snapshot_observed_at":"2026-08-06T12:33:59.046454Z","submitted_at":"2025-07-29T10:07:24Z","title":"The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM"},"reference_resolution":{"displayed":100,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":95,"verified_exact":5,"verified_fuzzy":0},"total_outbound_references":241},"refusal":"A citation records a reference. It does not transfer a finding from one paper to another.","schema":"pith.paper-citation-record.v1","standing_sources":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"thesis":"As of 7 August 2026, this Paper Citation Record lists 100 of 241 outbound references and 1 inbound Pith citation observation for arXiv:2507.21649."}