Papers › Target before Shooting: Accurate Anomaly Detection and Localization under One...

Target before Shooting: Accurate Anomaly Detection and Localization under One Millisecond via Cascade Patch Retrieval

13 Aug 2023arXiv:2308.06748archive 2025-07-28

Hanxi Li, Jianfei Hu, Bo Li, Hao Chen, Yongbin Zheng, Chunhua Shen

In this work, by re-examining the "matching" nature of Anomaly Detection (AD), we propose a new AD framework that simultaneously enjoys new records of AD accuracy and dramatically high running speed. In this framework, the anomaly detection problem is solved via a cascade patch retrieval procedure that retrieves the nearest neighbors for each test image patch in a coarse-to-fine fashion. Given a test sample, the top-K most similar training images are first selected based on a robust histogram matching process. Secondly, the nearest neighbor of each test patch is retrieved over the similar geometrical locations on those "global nearest neighbors", by using a carefully trained local metric. Finally, the anomaly score of each test image patch is calculated based on the distance to its "local nearest neighbor" and the "non-background" probability. The proposed method is termed "Cascade Patch Retrieval" (CPR) in this work. Different from the conventional patch-matching-based AD algorithms, CPR selects proper "targets" (reference images and locations) before "shooting" (patch-matching). On the well-acknowledged MVTec AD, BTAD and MVTec-3D AD datasets, the proposed algorithm consistently outperforms all the comparing SOTA methods by remarkable margins, measured by various AD metrics. Furthermore, CPR is extremely efficient. It runs at the speed of 113 FPS with the standard setting while its simplified version only requires less than 1 ms to process an image at the cost of a trivial accuracy drop. The code of CPR is available at https://github.com/flyinghu123/CPR.

PaperPDFCodeCode Syntology ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2308.06748")

Code

Syntology Ran 6 of 11 code samples harvested from 1 repository linked to this paper; 5 have no recorded run. Of those that ran: 6 ran with no contract checked.

By repository: official repository: 11 samples from 1 repository, 6 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

flyinghu123/cpr officialmentioned in paperpytorchMIT report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

11 samples harvested; 6 ran; 0 honoured the contract we drafted; 5 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

6ran
5unverified

Licence: 0 of the 11 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from flyinghu123/cpr. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: 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. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

compute_image_auc flyinghu123/cpr/metrics.py official repository ran fingerprinted MIT (permissive) · 9904f18fa913eb37 · report
entropy_pytorch flyinghu123/cpr/models/grb.py official repository ran fingerprinted MIT (permissive) · afd1fd4c67e83001 · report
fix_seeds flyinghu123/cpr/utils.py official repository ran MIT (permissive) · 0c3e0e613af57ee1 · report
get_grb flyinghu123/cpr/models/grb.py official repository ran MIT (permissive) · c05bbe23123f3bf7 · report
pixel_params_verify flyinghu123/cpr/metrics.py official repository ran MIT (permissive) · fc87506f23fa4651 · report
save_dependencies_files flyinghu123/cpr/utils.py official repository ran MIT (permissive) · 3eec1fa82531e363 · report
compute_image_auc_torch flyinghu123/cpr/metrics.py official repository unverified MIT (permissive) · aa3eaa45e0fc06c6 · report
generate_fractal_noise_2d flyinghu123/cpr/libs/perlin.py official repository unverified MIT (permissive) · 293a2b1189b747fd · report
generate_perlin_noise_2d flyinghu123/cpr/libs/perlin.py official repository unverified MIT (permissive) · ed9e127befc02046 · report
get_feb flyinghu123/cpr/models/feb.py official repository unverified MIT (permissive) · c0449268a49c4f3d · report
lerp_np flyinghu123/cpr/libs/perlin.py official repository unverified MIT (permissive) · 4de91a4627a51ea2 · report

Tasks

Anomaly DetectionSupervised Anomaly Detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Anomaly Detection BTAD CPR Detection AUROC 94.8 #9 of 15 Archive leaderboard report
Anomaly Detection BTAD CPR Segmentation AP 70.3 #9 of 15 Archive leaderboard report
Anomaly Detection BTAD CPR Segmentation AUPRO 85.1 #9 of 15 Archive leaderboard report
Anomaly Detection BTAD CPR Segmentation AUROC 98.4 #9 of 15 Archive leaderboard report
Anomaly Detection MVTec 3D-AD (RGB) CPR Detection AUROC 88.5 #2 of 2 Archive leaderboard report
Anomaly Detection MVTec 3D-AD (RGB) CPR Segmentation AP 57.8 #2 of 2 Archive leaderboard report
Anomaly Detection MVTec 3D-AD (RGB) CPR Segmentation AUPRO 96.9 #2 of 2 Archive leaderboard report
Anomaly Detection MVTec 3D-AD (RGB) CPR Segmentation AUROC 99.1 #2 of 2 Archive leaderboard report
Anomaly Detection MVTec AD CPR Detection AUROC 99.7 #12 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR FPS 113 #12 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR Segmentation AP 82.7 #12 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR Segmentation AUPRO 97.8 #12 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR Segmentation AUROC 99.2 #12 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR-fast Detection AUROC 99.7 #13 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR-fast FPS 245 #13 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR-fast Segmentation AP 82.3 #13 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR-fast Segmentation AUPRO 97.7 #13 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR-fast Segmentation AUROC 99.2 #13 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR-faster Detection AUROC 99.4 #34 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR-faster FPS 478 #34 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR-faster Segmentation AP 80.6 #34 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR-faster Segmentation AUPRO 97.3 #34 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR-faster Segmentation AUROC 99.0 #34 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR-faster(TensorRT) FPS 1016 #146 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR-fast(TensorRT) FPS 362 #147 of 148 Archive leaderboard report
Anomaly Detection MVTec AD CPR(TensorRT) FPS 130 #148 of 148 Archive leaderboard report
Supervised Anomaly Detection BTAD CPR Detection AUROC 98.3 #1 of 2 Archive leaderboard report
Supervised Anomaly Detection BTAD CPR Segmentation AP 84.0 #1 of 2 Archive leaderboard report
Supervised Anomaly Detection BTAD CPR Segmentation AUPRO 91.4 #1 of 2 Archive leaderboard report
Supervised Anomaly Detection BTAD CPR Segmentation AUROC 99.1 #1 of 2 Archive leaderboard report
Supervised Anomaly Detection MVTec AD CPR Detection AUROC 99.7 #2 of 8 Archive leaderboard report
Supervised Anomaly Detection MVTec AD CPR Segmentation AP 86.0 #2 of 8 Archive leaderboard report
Supervised Anomaly Detection MVTec AD CPR Segmentation AUPRO 98.3 #2 of 8 Archive leaderboard report
Supervised Anomaly Detection MVTec AD CPR Segmentation AUROC 99.6 #2 of 8 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

1x1 ConvolutionAuxiliary ClassifierAverage PoolingBatch NormalizationContrastive LearningConvolutionDense ConnectionsDropoutGlobal Average PoolingInception ModuleInception v2Kaiming InitializationLDAMax PoolingReLUResidual ConnectionSoftmaxWide Residual BlockWideResNet

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections