{"url":"/sota/anomaly-detection-on-btad","task":{"name":"Anomaly Detection","url":"/task/anomaly-detection","note":null},"dataset":{"name":"BTAD","url":"/dataset/btad"},"category":"Computer Vision","categories":["Computer Vision","Graphs","Methodology","Miscellaneous"],"category_note":null,"description":"**Anomaly Detection** is a binary classification identifying unusual or unexpected patterns in a dataset, which deviate significantly from the majority of the data. The goal of anomaly detection is to identify such anomalies, which could represent errors, fraud, or other types of unusual events, and flag them for further investigation.\r\n\r\n[Image source]: [GAN-based Anomaly Detection in Imbalance Problems](https://paperswithcode.com/paper/gan-based-anomaly-detection-in-imbalance)","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":["Detection AUROC","Segmentation AUROC","Segmentation AP","Segmentation AUPRO"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Detection AUROC":null,"Segmentation AUROC":null,"Segmentation AP":"higher","Segmentation AUPRO":null}},"counts":{"rows":15,"rows_with_code":12,"rows_with_paper_page":15,"rows_dated":15,"rows_using_additional_data":8},"rows":[{"rank_in_archive_order":1,"model":"UniNet","metrics":{"Detection AUROC":"97.73","Segmentation AUPRO":"80.01","Segmentation AUROC":"97.70"},"uses_additional_data":false,"paper_date":"2025-02-28","paper":"/paper/uninet-a-contrastive-learning-guided-unified","paper_url":"https://pangdatangtt.github.io/#:~:text=guided%20anomaly%20discrimination.-,Abstract,-Anomaly%20detection%20(AD","paper_title":"UniNet: A Contrastive Learning-guided Unified Framework with Feature Selection for Anomaly Detection","code":"https://github.com/pangdatangtt/UniNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"MuSc (zero-shot)","metrics":{"Detection AUROC":"96.16","Segmentation AP":"57.15","Segmentation AUPRO":"83.43","Segmentation AUROC":"97.35"},"uses_additional_data":false,"paper_date":"2024-01-30","paper":"/paper/musc-zero-shot-industrial-anomaly","paper_url":"https://arxiv.org/abs/2401.16753v1","paper_title":"MuSc: Zero-Shot Industrial Anomaly Classification and Segmentation with Mutual Scoring of the Unlabeled Images","code":"https://github.com/xrli-U/MuSc","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":3,"model":"RealNet","metrics":{"Detection AUROC":"96.1","Segmentation AUROC":"97.9"},"uses_additional_data":true,"paper_date":"2024-03-09","paper":"/paper/realnet-a-feature-selection-network-with","paper_url":"https://arxiv.org/abs/2403.05897v1","paper_title":"RealNet: A Feature Selection Network with Realistic Synthetic Anomaly for Anomaly Detection","code":"https://github.com/cnulab/realnet","n_code_links":1,"syntology":{"n_ran":9,"n_unverified":1,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":4,"model":"AD-CLSCNFs","metrics":{"Detection AUROC":"95.93","Segmentation AP":"55.86","Segmentation AUPRO":"72.77","Segmentation AUROC":"97.13"},"uses_additional_data":false,"paper_date":"2024-05-28","paper":"/paper/anomaly-detection-using-normalizing-flow","paper_url":"https://ieeexplore.ieee.org/document/10540095","paper_title":"Anomaly Detection Using Normalizing Flow-Based Density Estimation and Synthetic Defect Classification","code":"https://github.com/seungmi-oh/AD-CLSCNFs","n_code_links":1,"syntology":null},{"rank_in_archive_order":5,"model":"PyramidFlow (Res18)","metrics":{"Detection AUROC":"95.8","Segmentation AUROC":"97.7"},"uses_additional_data":true,"paper_date":"2023-03-05","paper":"/paper/pyramidflow-high-resolution-defect","paper_url":"https://arxiv.org/abs/2303.02595v1","paper_title":"PyramidFlow: High-Resolution Defect Contrastive Localization using Pyramid Normalizing Flow","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":6,"model":"ReConPatch WRN-50","metrics":{"Detection AUROC":"95.8","Segmentation AUPRO":"97.5"},"uses_additional_data":false,"paper_date":"2023-05-26","paper":"/paper/reconpatch-contrastive-patch-representation","paper_url":"https://arxiv.org/abs/2305.16713v3","paper_title":"ReConPatch : Contrastive Patch Representation Learning for Industrial Anomaly Detection","code":"https://github.com/travishsu/ReConPatch-TF","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":7,"model":"Reverse Distillation ++","metrics":{"Detection AUROC":"95.63","Segmentation AUROC":"97.43"},"uses_additional_data":false,"paper_date":"2023-01-01","paper":"/paper/revisiting-reverse-distillation-for-anomaly","paper_url":"http://openaccess.thecvf.com//content/CVPR2023/html/Tien_Revisiting_Reverse_Distillation_for_Anomaly_Detection_CVPR_2023_paper.html","paper_title":"Revisiting Reverse Distillation for Anomaly Detection","code":"https://github.com/tientrandinh/revisiting-reverse-distillation","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"D3AD","metrics":{"Detection AUROC":"95.2","Segmentation AUPRO":"83.2"},"uses_additional_data":false,"paper_date":"2024-01-09","paper":"/paper/d3ad-dynamic-denoising-diffusion","paper_url":"https://arxiv.org/abs/2401.04463v2","paper_title":"Dynamic Addition of Noise in a Diffusion Model for Anomaly Detection","code":"https://github.com/JustinTebbe/D3AD","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"CPR","metrics":{"Detection AUROC":"94.8","Segmentation AP":"70.3","Segmentation AUPRO":"85.1","Segmentation AUROC":"98.4"},"uses_additional_data":false,"paper_date":"2023-08-13","paper":"/paper/target-before-shooting-accurate-anomaly","paper_url":"https://arxiv.org/abs/2308.06748v1","paper_title":"Target before Shooting: Accurate Anomaly Detection and Localization under One Millisecond via Cascade Patch Retrieval","code":"https://github.com/flyinghu123/cpr","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":5,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":10,"model":"WeakREST-Un","metrics":{"Detection AUROC":"94.4","Segmentation AP":"63.1","Segmentation AUPRO":"84.9","Segmentation AUROC":"98.7"},"uses_additional_data":true,"paper_date":"2023-06-06","paper":"/paper/efficient-anomaly-detection-with-budget","paper_url":"https://arxiv.org/abs/2306.03492v6","paper_title":"Industrial Anomaly Detection and Localization Using Weakly-Supervised Residual Transformers","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":11,"model":"URD","metrics":{"Detection AUROC":"93.9","Segmentation AP":"65.2","Segmentation AUPRO":"78.5","Segmentation AUROC":"98.1"},"uses_additional_data":true,"paper_date":"2024-12-10","paper":"/paper/unlocking-the-potential-of-reverse","paper_url":"https://arxiv.org/abs/2412.07579v1","paper_title":"Unlocking the Potential of Reverse Distillation for Anomaly Detection","code":"https://github.com/hito2448/urd","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"FastFlow+AltUB","metrics":{"Segmentation AUROC":"98.2"},"uses_additional_data":true,"paper_date":"2022-10-26","paper":"/paper/altub-alternating-training-method-to-update","paper_url":"https://arxiv.org/abs/2210.14913v1","paper_title":"AltUB: Alternating Training Method to Update Base Distribution of Normalizing Flow for Anomaly Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":13,"model":"PNI","metrics":{"Segmentation AUROC":"97.8"},"uses_additional_data":true,"paper_date":"2022-11-22","paper":"/paper/image-anomaly-detection-and-localization-with","paper_url":"https://arxiv.org/abs/2211.12634v3","paper_title":"PNI : Industrial Anomaly Detection using Position and Neighborhood Information","code":"https://github.com/wogur110/PNI_anomaly_detection","n_code_links":1,"syntology":null},{"rank_in_archive_order":14,"model":"PatchSVDD","metrics":{"Segmentation AUROC":"93.1"},"uses_additional_data":true,"paper_date":"2020-06-29","paper":"/paper/patch-svdd-patch-level-svdd-for-anomaly","paper_url":"https://arxiv.org/abs/2006.16067v2","paper_title":"Patch SVDD: Patch-level SVDD for Anomaly Detection and Segmentation","code":"https://github.com/nuclearboy95/Anomaly-Detection-PatchSVDD-PyTorch","n_code_links":3,"syntology":null},{"rank_in_archive_order":15,"model":"VT-ADL","metrics":{"Segmentation AUROC":"81.8"},"uses_additional_data":true,"paper_date":"2021-04-20","paper":"/paper/vt-adl-a-vision-transformer-network-for-image","paper_url":"https://arxiv.org/abs/2104.10036v1","paper_title":"VT-ADL: A Vision Transformer Network for Image Anomaly Detection and Localization","code":"https://github.com/pankajmishra000/VT-ADL","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":12,"n_samples":12,"n_pointer_only_licence":0}}],"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 6,264 of the 9,581 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":9581,"papers_checked":6264,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":3316},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-24T18:15:14+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":5,"rows_with_any_sample_ran":3,"distinct_papers_with_graph_line":5,"distinct_papers_with_any_sample_ran":3,"samples_over_distinct_papers":{"n_ran":16,"n_unverified":19,"n_samples":35,"n_pointer_only_licence":0,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":16,"n_unverified":19,"n_samples":35,"n_pointer_only_licence":0,"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"}}}