{"url":"/sota/3d-anomaly-detection-on-real-3d-ad","task":{"name":"3D Anomaly Detection","url":"/task/3d-anomaly-detection","note":null},"dataset":{"name":"Real 3D-AD","url":"/dataset/real-3d-ad"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"3D-only Anomaly Detection.  Structures out of normal distribution are detected from the 3D-only point cloud.","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":["Mean Performance of P. and O. ","Point AUROC","Object AUROC"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Mean Performance of P. and O. ":null,"Point AUROC":null,"Object AUROC":null}},"counts":{"rows":19,"rows_with_code":14,"rows_with_paper_page":19,"rows_dated":19,"rows_using_additional_data":3},"rows":[{"rank_in_archive_order":1,"model":"DUS-Net","metrics":{"Mean Performance of P. and O. ":"0.821","Object AUROC":"0.795","Point AUROC":"0.847"},"uses_additional_data":false,"paper_date":"2025-07-05","paper":"/paper/taming-anomalies-with-down-up-sampling","paper_url":"https://arxiv.org/abs/2507.03903v1","paper_title":"Taming Anomalies with Down-Up Sampling Networks: Group Center Preserving Reconstruction for 3D Anomaly Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":2,"model":"MC4AD","metrics":{"Mean Performance of P. and O. ":"0.8115","Object AUROC":"0.786","Point AUROC":"0.837"},"uses_additional_data":true,"paper_date":"2025-05-09","paper":"/paper/examining-the-source-of-defects-from-a","paper_url":"https://arxiv.org/abs/2505.05901v2","paper_title":"Examining the Source of Defects from a Mechanical Perspective for 3D Anomaly Detection","code":"https://github.com/hzzzzzhappy/mc4ad","n_code_links":1,"syntology":null},{"rank_in_archive_order":3,"model":"ISMP","metrics":{"Mean Performance of P. and O. ":"0.7965","Object AUROC":"0.757","Point AUROC":"0.836"},"uses_additional_data":false,"paper_date":"2024-12-18","paper":"/paper/look-inside-for-more-internal-spatial","paper_url":"https://arxiv.org/abs/2412.13461v2","paper_title":"Look Inside for More: Internal Spatial Modality Perception for 3D Anomaly Detection","code":"https://github.com/M-3LAB/Look-Inside-for-More","n_code_links":1,"syntology":null},{"rank_in_archive_order":4,"model":"MC3D-AD","metrics":{"Mean Performance of P. and O. ":"0.775","Object AUROC":"0.782","Point AUROC":"0.768"},"uses_additional_data":false,"paper_date":"2025-05-04","paper":"/paper/mc3d-ad-a-unified-geometry-aware","paper_url":"https://arxiv.org/abs/2505.01969v1","paper_title":"MC3D-AD: A Unified Geometry-aware Reconstruction Model for Multi-category 3D Anomaly Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":5,"model":"PASDF","metrics":{"Mean Performance of P. and O. ":"0.7735","Object AUROC":"0.802","Point AUROC":"0.745"},"uses_additional_data":false,"paper_date":"2025-05-30","paper":"/paper/bridging-3d-anomaly-localization-and-repair","paper_url":"https://arxiv.org/abs/2505.24431v1","paper_title":"Bridging 3D Anomaly Localization and Repair via High-Quality Continuous Geometric Representation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":6,"model":"PO3AD","metrics":{"Mean Performance of P. and O. ":"0.765","Object AUROC":"0.765"},"uses_additional_data":false,"paper_date":"2024-12-17","paper":"/paper/po3ad-predicting-point-offsets-toward-better","paper_url":"https://arxiv.org/abs/2412.12617v1","paper_title":"PO3AD: Predicting Point Offsets toward Better 3D Point Cloud Anomaly Detection","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":7,"model":"Group3AD","metrics":{"Mean Performance of P. and O. ":"0.743","Object AUROC":"0.751","Point AUROC":"0.735"},"uses_additional_data":false,"paper_date":"2024-08-08","paper":"/paper/towards-high-resolution-3d-anomaly-detection","paper_url":"https://arxiv.org/abs/2408.04604v1","paper_title":"Towards High-resolution 3D Anomaly Detection via Group-Level Feature Contrastive Learning","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":8,"model":"PointAD","metrics":{"Mean Performance of P. and O. ":"0.7375","Object AUROC":"0.759","Point AUROC":"0.716"},"uses_additional_data":false,"paper_date":"2024-10-01","paper":"/paper/pointad-comprehending-3d-anomalies-from","paper_url":"https://arxiv.org/abs/2410.00320v4","paper_title":"PointAD: Comprehending 3D Anomalies from Points and Pixels for Zero-shot 3D Anomaly Detection","code":"https://github.com/zqhang/pointad","n_code_links":1,"syntology":{"n_ran":17,"n_unverified":9,"n_samples":26,"n_pointer_only_licence":23}},{"rank_in_archive_order":9,"model":"IMRNet","metrics":{"Mean Performance of P. and O. ":"0.725","Object AUROC":"0.725"},"uses_additional_data":true,"paper_date":"2023-11-25","paper":"/paper/towards-scalable-3d-anomaly-detection-and","paper_url":"https://arxiv.org/abs/2311.14897v3","paper_title":"Towards Scalable 3D Anomaly Detection and Localization: A Benchmark via 3D Anomaly Synthesis and A Self-Supervised Learning Network","code":"https://github.com/chopper-233/anomaly-shapenet","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"GLFM","metrics":{"Mean Performance of P. and O. ":"0.715","Object AUROC":"0.532","Point AUROC":"0.898"},"uses_additional_data":false,"paper_date":"2025-02-21","paper":"/paper/boosting-global-local-feature-matching-via","paper_url":"https://ieeexplore.ieee.org/document/10898004","paper_title":"Boosting Global-Local Feature Matching via Anomaly Synthesis for Multi-Class Point Cloud Anomaly Detection","code":"https://github.com/hustCYQ/GLFM-Multi-class-3DAD","n_code_links":1,"syntology":null},{"rank_in_archive_order":11,"model":"Reg3D-AD","metrics":{"Mean Performance of P. and O. ":"0.702","Object AUROC":"0.704","Point AUROC":"0.700"},"uses_additional_data":true,"paper_date":"2023-09-23","paper":"/paper/real3d-ad-a-dataset-of-point-cloud-anomaly-1","paper_url":"https://arxiv.org/abs/2309.13226v3","paper_title":"Real3D-AD: A Dataset of Point Cloud Anomaly Detection","code":"https://github.com/m-3lab/real3d-ad","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":3,"n_samples":5,"n_pointer_only_licence":5}},{"rank_in_archive_order":12,"model":"PatchCore (FPFH+Raw)","metrics":{"Mean Performance of P. and O. ":"0.687","Object AUROC":"0.682","Point AUROC":"0.692"},"uses_additional_data":false,"paper_date":"2021-06-15","paper":"/paper/towards-total-recall-in-industrial-anomaly","paper_url":"https://arxiv.org/abs/2106.08265v2","paper_title":"Towards Total Recall in Industrial Anomaly Detection","code":"https://github.com/openvinotoolkit/anomalib","n_code_links":18,"syntology":{"n_ran":5,"n_unverified":31,"n_samples":36,"n_pointer_only_licence":0}},{"rank_in_archive_order":13,"model":"BTF (Raw)","metrics":{"Mean Performance of P. and O. ":"0.6785","Object AUROC":"0.635","Point AUROC":"0.722"},"uses_additional_data":false,"paper_date":"2022-03-10","paper":"/paper/an-empirical-investigation-of-3d-anomaly","paper_url":"https://arxiv.org/abs/2203.05550v3","paper_title":"Back to the Feature: Classical 3D Features are (Almost) All You Need for 3D Anomaly Detection","code":"https://github.com/eliahuhorwitz/3D-ADS","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":8,"n_samples":9,"n_pointer_only_licence":1}},{"rank_in_archive_order":14,"model":"R3D-AD","metrics":{"Mean Performance of P. and O. ":"0.663","Object AUROC":"0.734","Point AUROC":"0.592"},"uses_additional_data":false,"paper_date":"2024-07-15","paper":"/paper/r3d-ad-reconstruction-via-diffusion-for-3d","paper_url":"https://arxiv.org/abs/2407.10862v1","paper_title":"R3D-AD: Reconstruction via Diffusion for 3D Anomaly Detection","code":"https://github.com/zhouzheyuan/r3d-ad","n_code_links":1,"syntology":null},{"rank_in_archive_order":15,"model":"PatchCore (PointMAE)","metrics":{"Mean Performance of P. and O. ":"0.614","Object AUROC":"0.594","Point AUROC":"0.634"},"uses_additional_data":false,"paper_date":"2021-06-15","paper":"/paper/towards-total-recall-in-industrial-anomaly","paper_url":"https://arxiv.org/abs/2106.08265v2","paper_title":"Towards Total Recall in Industrial Anomaly Detection","code":"https://github.com/openvinotoolkit/anomalib","n_code_links":18,"syntology":{"n_ran":5,"n_unverified":31,"n_samples":36,"n_pointer_only_licence":0}},{"rank_in_archive_order":16,"model":"M3DM (PointMAE)","metrics":{"Mean Performance of P. and O. ":"0.5945","Object AUROC":"0.552","Point AUROC":"0.637"},"uses_additional_data":false,"paper_date":"2023-03-01","paper":"/paper/multimodal-industrial-anomaly-detection-via","paper_url":"https://arxiv.org/abs/2303.00601v2","paper_title":"Multimodal Industrial Anomaly Detection via Hybrid Fusion","code":"https://github.com/nomewang/m3dm","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":6,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":17,"model":"PatchCore (FPFH)","metrics":{"Mean Performance of P. and O. ":"0.5925","Object AUROC":"0.593","Point AUROC":"0.592"},"uses_additional_data":false,"paper_date":"2021-06-15","paper":"/paper/towards-total-recall-in-industrial-anomaly","paper_url":"https://arxiv.org/abs/2106.08265v2","paper_title":"Towards Total Recall in Industrial Anomaly Detection","code":"https://github.com/openvinotoolkit/anomalib","n_code_links":18,"syntology":{"n_ran":5,"n_unverified":31,"n_samples":36,"n_pointer_only_licence":0}},{"rank_in_archive_order":18,"model":"M3DM (PointBERT)","metrics":{"Mean Performance of P. and O. ":"0.587","Object AUROC":"0.538","Point AUROC":"0.636"},"uses_additional_data":false,"paper_date":"2023-03-01","paper":"/paper/multimodal-industrial-anomaly-detection-via","paper_url":"https://arxiv.org/abs/2303.00601v2","paper_title":"Multimodal Industrial Anomaly Detection via Hybrid Fusion","code":"https://github.com/nomewang/m3dm","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":6,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":19,"model":"BTF (FPFH)","metrics":{"Mean Performance of P. and O. ":"0.5845","Object AUROC":"0.603","Point AUROC":"0.566"},"uses_additional_data":false,"paper_date":"2022-03-10","paper":"/paper/an-empirical-investigation-of-3d-anomaly","paper_url":"https://arxiv.org/abs/2203.05550v3","paper_title":"Back to the Feature: Classical 3D Features are (Almost) All You Need for 3D Anomaly Detection","code":"https://github.com/eliahuhorwitz/3D-ADS","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":8,"n_samples":9,"n_pointer_only_licence":1}}],"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 6,795 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":6795,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":2785},"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":9,"rows_with_any_sample_ran":9,"distinct_papers_with_graph_line":5,"distinct_papers_with_any_sample_ran":5,"samples_over_distinct_papers":{"n_ran":28,"n_unverified":57,"n_samples":85,"n_pointer_only_licence":29,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":42,"n_unverified":133,"n_samples":175,"n_pointer_only_licence":30,"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"}}}