{"url":"/sota/anomaly-detection-in-surveillance-videos-on","task":{"name":"Anomaly Detection In Surveillance Videos","url":"/task/anomaly-detection-in-surveillance-videos","note":null},"dataset":{"name":"UCF-Crime","url":"/dataset/ucf-crime"},"category":"Computer Vision","categories":["Computer Vision","Graphs","Methodology","Miscellaneous"],"category_note":null,"description":"\"The goal of a practical anomaly detection system is to timely signal an activity that deviates normal patterns and identify the time window of the occurring anomaly. [It] can be considered as coarse level video understanding, which filters out anomalies from normal patterns.\" A critical task in video surveillance is detecting anomalous events such as  traffic accidents, crimes or illegal activities. Anomalous events rarely occur as compared to normal activities. Hence the application of this task is to \"alleviate the waste of labor and time, developing intelligent computer vision algorithms for automatic video anomaly detection\".\r\n\r\n(Credit: Real-world Anomaly Detection in Surveillance Videos)","description_from":"task","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["ROC AUC","Decidability","EER","AUC"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"ROC AUC":"higher","Decidability":null,"EER":null,"AUC":"higher"}},"counts":{"rows":21,"rows_with_code":16,"rows_with_paper_page":21,"rows_dated":21,"rows_using_additional_data":1},"rows":[{"rank_in_archive_order":1,"model":"STEAD-Base","metrics":{"ROC AUC":"91.34"},"uses_additional_data":false,"paper_date":"2025-03-11","paper":"/paper/stead-spatio-temporal-efficient-anomaly-1","paper_url":"https://arxiv.org/abs/2503.07942v1","paper_title":"STEAD: Spatio-Temporal Efficient Anomaly Detection for Time and Compute Sensitive Applications","code":"https://github.com/agao8/STEAD","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"MTFL (VST, finetuned on VADD)","metrics":{"ROC AUC":"89.78"},"uses_additional_data":true,"paper_date":"2024-10-08","paper":"/paper/mtfl-multi-timescale-feature-learning-for","paper_url":"https://arxiv.org/abs/2410.05900v1","paper_title":"MTFL: Multi-Timescale Feature Learning for Weakly-Supervised Anomaly Detection in Surveillance Videos","code":"https://github.com/erktkdg/MTFL","n_code_links":1,"syntology":null},{"rank_in_archive_order":3,"model":"STEAD-Fast","metrics":{"ROC AUC":"88.87"},"uses_additional_data":false,"paper_date":"2025-03-11","paper":"/paper/stead-spatio-temporal-efficient-anomaly-1","paper_url":"https://arxiv.org/abs/2503.07942v1","paper_title":"STEAD: Spatio-Temporal Efficient Anomaly Detection for Time and Compute Sensitive Applications","code":"https://github.com/agao8/STEAD","n_code_links":1,"syntology":null},{"rank_in_archive_order":4,"model":"BN-WVAD","metrics":{"ROC AUC":"87.24"},"uses_additional_data":false,"paper_date":"2023-11-26","paper":"/paper/batchnorm-based-weakly-supervised-video","paper_url":"https://arxiv.org/abs/2311.15367v1","paper_title":"BatchNorm-based Weakly Supervised Video Anomaly Detection","code":"https://github.com/cool-xuan/bn-wvad","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":2,"n_samples":5,"n_pointer_only_licence":5}},{"rank_in_archive_order":5,"model":"MTFL (VST)","metrics":{"ROC AUC":"87.16"},"uses_additional_data":false,"paper_date":"2024-10-08","paper":"/paper/mtfl-multi-timescale-feature-learning-for","paper_url":"https://arxiv.org/abs/2410.05900v1","paper_title":"MTFL: Multi-Timescale Feature Learning for Weakly-Supervised Anomaly Detection in Surveillance Videos","code":"https://github.com/erktkdg/MTFL","n_code_links":1,"syntology":null},{"rank_in_archive_order":6,"model":"ProDisc-VAD","metrics":{"ROC AUC":"87.12"},"uses_additional_data":false,"paper_date":"2025-05-04","paper":"/paper/prodisc-vad-an-efficient-system-for-weakly","paper_url":"https://arxiv.org/abs/2505.02179v1","paper_title":"ProDisc-VAD: An Efficient System for Weakly-Supervised Anomaly Detection in Video Surveillance Applications","code":"https://github.com/modadundun/ProDisc-VAD","n_code_links":1,"syntology":null},{"rank_in_archive_order":7,"model":"MGFN","metrics":{"ROC AUC":"86.98"},"uses_additional_data":false,"paper_date":"2022-11-28","paper":"/paper/mgfn-magnitude-contrastive-glance-and-focus","paper_url":"https://arxiv.org/abs/2211.15098v1","paper_title":"MGFN: Magnitude-Contrastive Glance-and-Focus Network for Weakly-Supervised Video Anomaly Detection","code":"https://github.com/carolchenyx/mgfn","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"PEL","metrics":{"ROC AUC":"86.76"},"uses_additional_data":false,"paper_date":"2023-06-26","paper":"/paper/learning-prompt-enhanced-context-features-for","paper_url":"https://arxiv.org/abs/2306.14451v2","paper_title":"Learning Prompt-Enhanced Context Features for Weakly-Supervised Video Anomaly Detection","code":"https://github.com/yujiangpu20/pel4vad","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":9,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":9,"model":"S3R","metrics":{"ROC AUC":"85.99"},"uses_additional_data":false,"paper_date":"2022-10-23","paper":"/paper/self-supervised-sparse-representation-for","paper_url":"https://www.ecva.net/papers/eccv_2022/papers_ECCV/papers/136730727.pdf","paper_title":"Self-supervised Sparse Representation for Video Anomaly Detection","code":"https://github.com/louisYen/S3R","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"WSAL","metrics":{"ROC AUC":"85.38"},"uses_additional_data":false,"paper_date":"2020-08-20","paper":"/paper/localizing-anomalies-from-weakly-labeled","paper_url":"https://arxiv.org/abs/2008.08944v3","paper_title":"Localizing Anomalies from Weakly-Labeled Videos","code":"https://github.com/ktr-hubrt/WSAL","n_code_links":1,"syntology":null},{"rank_in_archive_order":11,"model":"CR-UNet","metrics":{"ROC AUC":"85.24"},"uses_additional_data":false,"paper_date":"2023-04-11","paper":"/paper/contrastive-regularized-u-net-for-video","paper_url":"https://ieeexplore.ieee.org/abstract/document/10098744/","paper_title":"Contrastive-Regularized U-Net for Video Anomaly Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":12,"model":"Multi-stream Network with Late Fuzzy Fusion","metrics":{"ROC AUC":"84.48"},"uses_additional_data":false,"paper_date":"2022-03-27","paper":"/paper/a-multi-stream-deep-neural-network-with-late","paper_url":"https://doi.org/10.1016/j.eswa.2022.117030","paper_title":"A multi-stream deep neural network with late fuzzy fusion for real-world anomaly detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":13,"model":"RTFM","metrics":{"ROC AUC":"84.03"},"uses_additional_data":false,"paper_date":"2021-01-25","paper":"/paper/weakly-supervised-video-anomaly-detection","paper_url":"https://arxiv.org/abs/2101.10030v3","paper_title":"Weakly-supervised Video Anomaly Detection with Robust Temporal Feature Magnitude Learning","code":"https://github.com/tianyu0207/RTFM","n_code_links":3,"syntology":{"n_ran":4,"n_unverified":1,"n_samples":5,"n_pointer_only_licence":5}},{"rank_in_archive_order":14,"model":"MIST","metrics":{"ROC AUC":"82.30"},"uses_additional_data":false,"paper_date":"2021-04-04","paper":"/paper/mist-multiple-instance-self-training","paper_url":"https://arxiv.org/abs/2104.01633v1","paper_title":"MIST: Multiple Instance Self-Training Framework for Video Anomaly Detection","code":"https://github.com/fjchange/MIST_VAD","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":15,"model":"DMRMs","metrics":{"Decidability":"-","EER":"-","ROC AUC":"81.91"},"uses_additional_data":false,"paper_date":"2021-02-02","paper":"/paper/anomalous-event-recognition-in-videos-based","paper_url":"https://doi.org/10.3390/app11031344","paper_title":"Anomalous Event Recognition in Videos Based on Joint Learningof Motion and Appearance with Multiple Ranking Measures","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":16,"model":"Multiple-Instance-Based-Video-Anomaly-Detection-Using-Deep-Temporal-Encoding-Decoding","metrics":{"Decidability":"-","EER":"-","ROC AUC":"80.10"},"uses_additional_data":false,"paper_date":"2020-07-03","paper":"/paper/multiple-instance-based-video-anomaly","paper_url":"https://arxiv.org/abs/2007.01548v2","paper_title":"Multiple Instance-Based Video Anomaly Detection using Deep Temporal Encoding-Decoding","code":"https://github.com/AmmarKamoona/Multiple-Instance-Based-Video-Anomaly-Detection-Using-Deep-Temporal-Encoding-Decoding","n_code_links":1,"syntology":null},{"rank_in_archive_order":17,"model":"MULDE-frame-centric-micro-one-class-classification","metrics":{"ROC AUC":"78.5%"},"uses_additional_data":false,"paper_date":"2024-03-21","paper":"/paper/mulde-multiscale-log-density-estimation-via","paper_url":"https://arxiv.org/abs/2403.14497v1","paper_title":"MULDE: Multiscale Log-Density Estimation via Denoising Score Matching for Video Anomaly Detection","code":"https://github.com/jakubmicorek/MULDE-Multiscale-Log-Density-Estimation-via-Denoising-Score-Matching-for-Video-Anomaly-Detection","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":18,"model":"MILR","metrics":{"Decidability":"-","EER":"-","ROC AUC":"76.67"},"uses_additional_data":false,"paper_date":"2020-02-04","paper":"/paper/3d-resnet-with-ranking-loss-function-for","paper_url":"https://arxiv.org/abs/2002.01132v1","paper_title":"3D ResNet with Ranking Loss Function for Abnormal Activity Detection in Videos","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":19,"model":"GMM-based","metrics":{"Decidability":"0.885","EER":"0.302","ROC AUC":"75.90"},"uses_additional_data":false,"paper_date":"2020-07-23","paper":"/paper/weakly-and-partially-supervised-learning","paper_url":"https://www.researchgate.net/publication/350580836_Weakly_and_Partially_Supervised_Learning_Frameworks_for_Anomaly_Detection","paper_title":"Weakly and Partially Supervised Learning Frameworks for Anomaly Detection","code":"https://github.com/DegardinBruno/human_self_learning_anomaly","n_code_links":1,"syntology":null},{"rank_in_archive_order":20,"model":"Sultani et al.","metrics":{"Decidability":"0.613","EER":"0.353","ROC AUC":"75.41"},"uses_additional_data":false,"paper_date":"2018-01-12","paper":"/paper/real-world-anomaly-detection-in-surveillance","paper_url":"http://arxiv.org/abs/1801.04264v3","paper_title":"Real-world Anomaly Detection in Surveillance Videos","code":"https://github.com/WaqasSultani/AnomalyDetectionCVPR2018","n_code_links":9,"syntology":{"n_ran":4,"n_unverified":2,"n_samples":6,"n_pointer_only_licence":4}},{"rank_in_archive_order":21,"model":"DAKD (Weakly-supervised)","metrics":{"AUC":"88.34"},"uses_additional_data":false,"paper_date":"2024-06-05","paper":"/paper/distilling-aggregated-knowledge-for-weakly","paper_url":"https://arxiv.org/abs/2406.02831v2","paper_title":"Distilling Aggregated Knowledge for Weakly-Supervised Video Anomaly Detection","code":null,"n_code_links":0,"syntology":null}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 6,885 of the 9,623 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9623,"papers_checked":6885,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":2737},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-25T09:33:49+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":6,"rows_with_any_sample_ran":5,"distinct_papers_with_graph_line":6,"distinct_papers_with_any_sample_ran":5,"samples_over_distinct_papers":{"n_ran":13,"n_unverified":14,"n_samples":27,"n_pointer_only_licence":16,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":13,"n_unverified":14,"n_samples":27,"n_pointer_only_licence":16,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}