{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/an-empirical-investigation-of-3d-anomaly","title":"Back to the Feature: Classical 3D Features are (Almost) All You Need for 3D Anomaly Detection","arxiv_id":"2203.05550","date":"2022-03-10","proceeding":null,"authors":["Eliahu Horwitz","Yedid Hoshen"],"abstract":"Despite significant advances in image anomaly detection and segmentation, few methods use 3D information. We utilize a recently introduced 3D anomaly detection dataset to evaluate whether or not using 3D information is a lost opportunity. First, we present a surprising finding: standard color-only methods outperform all current methods that are explicitly designed to exploit 3D information. This is counter-intuitive as even a simple inspection of the dataset shows that color-only methods are insufficient for images containing geometric anomalies. This motivates the question: how can anomaly detection methods effectively use 3D information? We investigate a range of shape representations including hand-crafted and deep-learning-based; we demonstrate that rotation invariance plays the leading role in the performance. We uncover a simple 3D-only method that beats all recent approaches while not using deep learning, external pre-training datasets, or color information. As the 3D-only method cannot detect color and texture anomalies, we combine it with color-based features, significantly outperforming previous state-of-the-art. Our method, dubbed BTF (Back to the Feature) achieves pixel-wise ROCAUC: 99.3% and PRO: 96.4% on MVTec 3D-AD.","url_abs":"https://arxiv.org/abs/2203.05550v3","url_pdf":"https://arxiv.org/pdf/2203.05550v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"an-empirical-investigation-of-3d-anomaly","repo_url":"https://github.com/eliahuhorwitz/3D-ADS","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"3d-anomaly-detection","task_name":"3D Anomaly Detection"},{"task_slug":"3d-anomaly-detection-and-segmentation","task_name":"3D Anomaly Detection and Segmentation"},{"task_slug":"all","task_name":"All"},{"task_slug":"anomaly-detection","task_name":"Anomaly Detection"},{"task_slug":"depth-anomaly-detection-and-segmentation","task_name":"Depth Anomaly Detection and Segmentation"},{"task_slug":"rgb-3d-anomaly-detection-and-segmentation","task_name":"RGB+3D Anomaly Detection and Segmentation"},{"task_slug":"rgb-depth-anomaly-detection-and-segmentation","task_name":"RGB+Depth Anomaly Detection and Segmentation"}],"methods":[{"method_slug":"btf","method_name":"BTF"}],"datasets_introduced":[],"methods_introduced":[{"slug":"btf","name":"BTF","full_name":"Back to the Feature"}],"results":[{"leaderboard":"/sota/3d-anomaly-detection-on-anomaly-shapenet10","task":"3D Anomaly Detection","dataset":"Anomaly-ShapeNet10","model":"BTF (FPFH)","rank_in_archive_order":3,"of":7,"metrics":{"O-AUROC":" 0.632","P-AUROC":" 0.790"},"uses_additional_data":false},{"leaderboard":"/sota/3d-anomaly-detection-on-anomaly-shapenet10","task":"3D Anomaly Detection","dataset":"Anomaly-ShapeNet10","model":"BTF (Raw)","rank_in_archive_order":7,"of":7,"metrics":{"O-AUROC":"0.500","P-AUROC":"0.515"},"uses_additional_data":false},{"leaderboard":"/sota/3d-anomaly-detection-on-real-3d-ad","task":"3D Anomaly Detection","dataset":"Real 3D-AD","model":"BTF (Raw)","rank_in_archive_order":13,"of":19,"metrics":{"Mean Performance of P. and O. ":"0.6785","Object AUROC":"0.635","Point AUROC":"0.722"},"uses_additional_data":false},{"leaderboard":"/sota/3d-anomaly-detection-on-real-3d-ad","task":"3D Anomaly Detection","dataset":"Real 3D-AD","model":"BTF (FPFH)","rank_in_archive_order":19,"of":19,"metrics":{"Mean Performance of P. and O. ":"0.5845","Object AUROC":"0.603","Point AUROC":"0.566"},"uses_additional_data":false},{"leaderboard":"/sota/3d-anomaly-detection-and-segmentation-on","task":"3D Anomaly Detection and Segmentation","dataset":"MVTEC 3D-AD","model":"Back to the Feature:\nClassical 3D Features are (Almost) All You Need for 3D Anomaly Detection  (FPFH)","rank_in_archive_order":7,"of":11,"metrics":{"Detection AUROC":"0.782","Segmentation AUROC":"0.978"},"uses_additional_data":false},{"leaderboard":"/sota/depth-anomaly-detection-and-segmentation-on","task":"Depth Anomaly Detection and Segmentation","dataset":"MVTEC 3D-AD","model":"Back to the Feature:\nClassical 3D Features are (Almost) All You Need for 3D Anomaly Detection (SIFT)","rank_in_archive_order":5,"of":13,"metrics":{"Detection AUROC":"0.727","Segmentation AUPRO":"0.910","Segmentation AUROC":"0.974"},"uses_additional_data":false},{"leaderboard":"/sota/depth-anomaly-detection-and-segmentation-on","task":"Depth Anomaly Detection and Segmentation","dataset":"MVTEC 3D-AD","model":"Back to the Feature:\nClassical 3D Features are (Almost) All You Need for 3D Anomaly Detection  (HoG)","rank_in_archive_order":7,"of":13,"metrics":{"Detection AUROC":"0.559","Segmentation AUPRO":"0.771","Segmentation AUROC":"0.930"},"uses_additional_data":false},{"leaderboard":"/sota/depth-anomaly-detection-and-segmentation-on","task":"Depth Anomaly Detection and Segmentation","dataset":"MVTEC 3D-AD","model":"Back to the Feature:\nClassical 3D Features are (Almost) All You Need for 3D Anomaly Detection  (Depth iNet)","rank_in_archive_order":8,"of":13,"metrics":{"Detection AUROC":"0.675","Segmentation AUPRO":"0.755","Segmentation AUROC":"0.930"},"uses_additional_data":true},{"leaderboard":"/sota/depth-anomaly-detection-and-segmentation-on","task":"Depth Anomaly Detection and Segmentation","dataset":"MVTEC 3D-AD","model":"Back to the Feature:\nClassical 3D Features are (Almost) All You Need for 3D Anomaly Detection  (NSA)","rank_in_archive_order":9,"of":13,"metrics":{"Detection AUROC":"0.696","Segmentation AUPRO":"0.5572","Segmentation AUROC":"0.817"},"uses_additional_data":false},{"leaderboard":"/sota/depth-anomaly-detection-and-segmentation-on","task":"Depth Anomaly Detection and Segmentation","dataset":"MVTEC 3D-AD","model":"Back to the Feature:\nClassical 3D Features are (Almost) All You Need for 3D Anomaly Detection  (RaW)","rank_in_archive_order":10,"of":13,"metrics":{"Detection AUROC":"0.573","Segmentation AUPRO":"0.442","Segmentation AUROC":"0.771"},"uses_additional_data":false},{"leaderboard":"/sota/rgb-3d-anomaly-detection-and-segmentation-on","task":"RGB+3D Anomaly Detection and Segmentation","dataset":"MVTEC 3D-AD","model":"Back to the Feature:\nClassical 3D Features are (Almost) All You Need for 3D Anomaly Detection (BTF)","rank_in_archive_order":6,"of":9,"metrics":{"Detection AUCROC":"0.865","Segmentation AUCROC":"0.992","Segmentation AUPRO":"0.959"},"uses_additional_data":true}],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=2203.05550","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2203.05550"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/eliahuhorwitz/3D-ADS","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_fixture":1,"unverified":8},"by_repo_kind":{"official":{"samples":8,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":1,"samples":[{"code_sha256_prefix":"f776b0618055e2f4","entry":"get_edges_of_pc","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"f776b0618055e2f4"}},{"code_sha256_prefix":"18c9bfa7e94951e9","entry":"collect_anomaly_scores","repo":"eliahuhorwitz/3D-ADS","repo_kind":"official","path":"utils/au_pro_util.py","file_url":"https://github.com/eliahuhorwitz/3D-ADS/blob/HEAD/utils/au_pro_util.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"18c9bfa7e94951e9"}},{"code_sha256_prefix":"49b5692178958afe","entry":"compute_pro","repo":"eliahuhorwitz/3D-ADS","repo_kind":"official","path":"utils/au_pro_util.py","file_url":"https://github.com/eliahuhorwitz/3D-ADS/blob/HEAD/utils/au_pro_util.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"49b5692178958afe"}},{"code_sha256_prefix":"501cb82dcd37b5e7","entry":"get_sift_bin_ksize_stride_pad","repo":"eliahuhorwitz/3D-ADS","repo_kind":"official","path":"utils/DenseSIFTDescriptor.py","file_url":"https://github.com/eliahuhorwitz/3D-ADS/blob/HEAD/utils/DenseSIFTDescriptor.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"501cb82dcd37b5e7"}},{"code_sha256_prefix":"21dbe36d5d11746a","entry":"get_sift_pooling_kernel","repo":"eliahuhorwitz/3D-ADS","repo_kind":"official","path":"utils/DenseSIFTDescriptor.py","file_url":"https://github.com/eliahuhorwitz/3D-ADS/blob/HEAD/utils/DenseSIFTDescriptor.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"21dbe36d5d11746a"}},{"code_sha256_prefix":"65012063890a92f8","entry":"organized_pc_to_unorganized_pc","repo":"eliahuhorwitz/3D-ADS","repo_kind":"official","path":"utils/mvtec3d_util.py","file_url":"https://github.com/eliahuhorwitz/3D-ADS/blob/HEAD/utils/mvtec3d_util.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"65012063890a92f8"}},{"code_sha256_prefix":"45b5420cddbb1bde","entry":"read_tiff_organized_pc","repo":"eliahuhorwitz/3D-ADS","repo_kind":"official","path":"utils/mvtec3d_util.py","file_url":"https://github.com/eliahuhorwitz/3D-ADS/blob/HEAD/utils/mvtec3d_util.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"45b5420cddbb1bde"}},{"code_sha256_prefix":"f4f7ac4f1e370dd3","entry":"resize_organized_pc","repo":"eliahuhorwitz/3D-ADS","repo_kind":"official","path":"utils/mvtec3d_util.py","file_url":"https://github.com/eliahuhorwitz/3D-ADS/blob/HEAD/utils/mvtec3d_util.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f4f7ac4f1e370dd3"}},{"code_sha256_prefix":"9d72c1f1f0d6dea4","entry":"trapezoid","repo":"eliahuhorwitz/3D-ADS","repo_kind":"official","path":"utils/au_pro_util.py","file_url":"https://github.com/eliahuhorwitz/3D-ADS/blob/HEAD/utils/au_pro_util.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9d72c1f1f0d6dea4"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}