Anomaly Factory 3D: A Modular Framework for Diverse Pseudo-Anomaly Synthesis in Unsupervised 3D Anomaly Detection

Ali Balapour; Faraz Hach

arxiv: 2606.29181 · v1 · pith:XXIDFHDCnew · submitted 2026-06-28 · 💻 cs.CV · cs.AI

Anomaly Factory 3D: A Modular Framework for Diverse Pseudo-Anomaly Synthesis in Unsupervised 3D Anomaly Detection

Ali Balapour , Faraz Hach This is my paper

Pith reviewed 2026-06-30 07:51 UTC · model grok-4.3

classification 💻 cs.CV cs.AI

keywords 3D anomaly detectionpseudo-anomaly synthesispoint cloud deformationparametric modelunsupervised learningAnomalyShapeNetReal3D-AD

0 comments

The pith

A parametric deformation framework synthesizes diverse pseudo-anomalies from normal 3D point clouds to train better unsupervised anomaly detectors.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces Anomaly Factory 3D, a framework that generates synthetic anomalies by applying controlled deformations to normal point cloud data. This addresses the scarcity of anomalous samples in unsupervised 3D anomaly detection. The deformations use a center-conditioned model in local PCA frames with adjustable parameters for spatial effects and displacement directions. When integrated into existing detectors, it leads to improved performance on benchmark datasets like AnomalyShapeNet and Real3D-AD. The tool is modular and can work with various detection methods.

Core claim

AF3AD employs a center-conditioned parametric deformation model defined in local PCA frames, incorporating kernel-controlled spatial falloff, anisotropy, directional gating, and normal/tangential displacement fields. This model enables the creation of a broad set of geometric defect presets from normal data. Integration with offset-prediction and reconstruction-based detectors demonstrates transferability across paradigms and yields improvements in object- and point-level detection and localization on AnomalyShapeNet and Real3D-AD.

What carries the argument

Center-conditioned parametric deformation model in local PCA frames with kernel-controlled falloff, anisotropy, directional gating, and displacement fields.

Load-bearing premise

The parametric deformation presets produce pseudo-anomalies that are sufficiently diverse and representative of real geometric defects to yield measurable gains when used for training.

What would settle it

A direct comparison experiment showing no improvement or a drop in detection and localization metrics when detectors are trained with AF3AD-augmented data versus normal data alone on AnomalyShapeNet or Real3D-AD would falsify the central claim.

Figures

Figures reproduced from arXiv: 2606.29181 by Ali Balapour, Faraz Hach.

**Figure 1.** Figure 1: Overview of AF3AD pipeline for geometric pseudo-anomaly synthesis. From a non-anomalous point cloud and its surface normals, AF3AD selects a local segment and computes a PCA-based local frame. Within this segment, each point is transformed into local coordinates, evaluated through an anisotropic distance function, and displaced according to preset-specified gating, direction, and deformation parameters. Th… view at source ↗

**Figure 2.** Figure 2: Samples of anomaly prediction using offset prediction method trained with AF3AD. 6 Conclusion In this work, we proposed AF3AD, a modular framework for synthesizing diverse pseudo-anomalies to generate training signals for 3D anomaly detection. Our framework provides a systematic taxonomy of geometric deformations with explicit parametric control, operating independently of specific detection architecture… view at source ↗

read the original abstract

Detecting and localizing defects in 3D point clouds is challenging because abnormal samples are scarce and diverse, while training is often limited to normal data. We propose Anomaly Factory 3D (AF3AD), a modular framework that synthesizes diverse pseudo-anomalies from normal point clouds to expand the training data for unsupervised 3D anomaly detection methods that rely on pseudo-anomalies. AF3AD uses a center-conditioned parametric deformation model defined in local PCA frames, with kernel-controlled spatial falloff, anisotropy, directional gating, and normal/tangential displacement fields, enabling a broad set of geometric defect presets. We demonstrate its ease-of-use and effectiveness by integrating AF3AD with an offset-prediction detector and a reconstruction-based anomaly detection method, showing that AF3AD transfers across detection paradigms. Experiments on AnomalyShapeNet and Real3D-AD show consistent improvements in object- and point-level detection and localization, supported by ablations on preset groups and robustness under noise. AF3AD is designed as a standalone synthesis tool to facilitate adoption across different 3D anomaly detection paradigms. Code is available at github.com/vpc-ccg/AF3AD.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

AF3AD introduces a parametric deformation model in local PCA frames for pseudo-anomaly synthesis and shows gains when added to two detectors, but the claim that these outputs are representative of real defects rests on unshown distributional evidence.

read the letter

AF3AD is a modular synthesis tool that deforms normal 3D point clouds using center-conditioned parametric changes inside local PCA frames, with kernel falloff, anisotropy, directional gating, and separate normal/tangential fields to create preset geometric defects. The main new element is this specific combination of controls packaged as a standalone generator meant to plug into different unsupervised detectors.

The paper does a solid job showing the framework transfers: it improves both an offset-prediction detector and a reconstruction-based one on object-level and point-level metrics for AnomalyShapeNet and Real3D-AD. The preset-group ablations and noise-robustness checks give some practical information about which parts matter. Releasing code is a clear positive for anyone who wants to test the presets directly.

The soft spot is exactly the one in the stress-test note. The central assumption is that the parametric presets produce anomalies whose statistics are close enough to real geometric defects to drive the measured gains. The abstract and the reported results do not include direct comparisons such as curvature histograms or displacement spectra between the synthesized samples and held-out real anomalies. Without that, it remains possible that the improvements come from generic added variety rather than targeted defect simulation. The lack of error bars or dataset statistics in the abstract also makes it harder to judge how stable the gains are.

This paper is for researchers working on 3D anomaly detection in industrial point-cloud settings who need more training variety when only normal data is available. A reader already using pseudo-anomaly methods would get concrete implementation details and cross-method results.

It has enough of a distinct technique and experimental coverage to deserve peer review rather than a desk reject.

Referee Report

1 major / 2 minor

Summary. The paper proposes Anomaly Factory 3D (AF3AD), a modular framework for synthesizing diverse pseudo-anomalies from normal 3D point clouds via a center-conditioned parametric deformation model defined in local PCA frames. The model incorporates kernel-controlled spatial falloff, anisotropy, directional gating, and normal/tangential displacement fields to support a broad set of geometric defect presets. AF3AD is integrated with an offset-prediction detector and a reconstruction-based detector to demonstrate transfer across paradigms, with experiments on AnomalyShapeNet and Real3D-AD reporting consistent gains in object- and point-level detection and localization, supported by preset-group ablations and noise-robustness tests. The framework is positioned as a standalone synthesis tool, with code released at github.com/vpc-ccg/AF3AD.

Significance. If the central claim holds, AF3AD supplies a practical, reusable tool for expanding training data in unsupervised 3D anomaly detection without requiring real anomalous samples. The modular design and demonstrated transfer across detection paradigms (offset prediction and reconstruction) could facilitate broader adoption. The public code release supports reproducibility and community use.

major comments (1)

[Experiments on AnomalyShapeNet and Real3D-AD] Experimental evaluation on AnomalyShapeNet and Real3D-AD: the reported improvements and preset-group ablations are consistent with the method's utility, yet the paper supplies no direct distributional comparison (e.g., curvature histograms, displacement spectra, or coverage of the defect manifold) between the synthesized pseudo-anomalies and held-out real anomalies. This comparison is load-bearing for attributing gains to the specific parametric deformation presets rather than generic augmentation effects.

minor comments (2)

[Abstract] Abstract: quantitative details on the magnitude of improvements, error bars, and dataset statistics are absent, which reduces the ability to assess effect sizes from the summary alone.
[Method description] The description of the deformation model would benefit from explicit notation or a table listing the free parameters of the kernel falloff, anisotropy, and directional gating components to clarify the preset construction process.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive comment on our experimental evaluation. We address the point below and will incorporate revisions to strengthen the attribution of results.

read point-by-point responses

Referee: [Experiments on AnomalyShapeNet and Real3D-AD] Experimental evaluation on AnomalyShapeNet and Real3D-AD: the reported improvements and preset-group ablations are consistent with the method's utility, yet the paper supplies no direct distributional comparison (e.g., curvature histograms, displacement spectra, or coverage of the defect manifold) between the synthesized pseudo-anomalies and held-out real anomalies. This comparison is load-bearing for attributing gains to the specific parametric deformation presets rather than generic augmentation effects.

Authors: We agree that a direct distributional comparison would provide stronger evidence that performance gains arise from the specific parametric presets rather than generic augmentation. While the preset-group ablations already isolate the contribution of individual deformation components, we acknowledge the referee's point that this is insufficient to fully rule out generic effects. In the revised manuscript we will add the requested analyses, including curvature histograms, displacement spectra, and coverage metrics comparing synthesized pseudo-anomalies against held-out real anomalies from both AnomalyShapeNet and Real3D-AD. revision: yes

Circularity Check

0 steps flagged

No significant circularity; synthesis framework is independent of downstream results.

full rationale

The paper introduces AF3AD as a parametric deformation model for generating pseudo-anomalies from normal point clouds, with components like center-conditioned PCA frames, kernel falloff, and displacement fields. No equations, fitted parameters, or self-citations are shown that reduce any claimed prediction or uniqueness result to the inputs by construction. The method is explicitly positioned as a standalone tool whose effectiveness is evaluated via integration with separate detectors on external benchmarks (AnomalyShapeNet, Real3D-AD), with ablations on preset groups. This matches the default expectation of a non-circular methodological contribution.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review provides insufficient detail to enumerate specific free parameters or axioms; the deformation model likely involves multiple tunable controls whose values are not specified here.

pith-pipeline@v0.9.1-grok · 5749 in / 1044 out tokens · 34611 ms · 2026-06-30T07:51:45.536494+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

30 extracted references · 27 canonical work pages · 1 internal anchor

[1]

CoRRabs/2112.09045 (2021), https://arxiv.org/abs/2112.09045

Bergmann, P., Jin, X., Sattlegger, D., Steger, C.: The mvtec 3d-ad dataset for un- supervised 3d anomaly detection and localization. CoRRabs/2112.09045 (2021), https://arxiv.org/abs/2112.09045

work page arXiv 2021
[2]

Pattern Recognit

Cao, Y., Xu, X., Shen, W.: Complementary pseudo multimodal feature for point cloud anomaly detection. Pattern Recognit. 156, 110761 (2024). https://doi. org/10.1016/J.PATCOG.2024.110761 , https://doi.org/10.1016/j.patcog. 2024.110761

work page doi:10.1016/j.patcog.2024.110761 2024
[3]

arXiv preprint arXiv:2401.16402 (2024)

Cao, Y., Xu, X., Zhang, J., Cheng, Y., Huang, X., Pang, G., Shen, W.: A survey on visual anomaly detection: Challenge, approach, and prospect. CoRR abs/2401.16402 (2024). https://doi.org/10.48550/ARXIV.2401.16402, https://doi.org/10.48550/arXiv.2401.16402

work page doi:10.48550/arxiv.2401.16402 2024
[4]

ShapeNet: An Information-Rich 3D Model Repository

Chang, A.X., Funkhouser, T.A., Guibas, L.J., Hanrahan, P., Huang, Q., Li, Z., Savarese, S., Savva, M., Song, S., Su, H., Xiao, J., Yi, L., Yu, F.: Shapenet: An information-rich 3d model repository. CoRR abs/1512.03012 (2015), http:// arxiv.org/abs/1512.03012

work page internal anchor Pith review Pith/arXiv arXiv 2015
[5]

IEEE Trans Autom

Cheng, Y., Cao, Y., Wang, D., Shen, W., Li, W.: Boosting global-local feature matching via anomaly synthesis for multi-class point cloud anomaly detection. IEEE Trans Autom. Sci. Eng. 22, 12560–12571 (2025).https://doi.org/10. 1109/TASE.2025.3544462, https://doi.org/10.1109/TASE.2025.3544462

work page doi:10.1109/tase.2025.3544462 2025
[6]

In: Koenig, S., Jenkins, C., Taylor, M.E

Cheng, Y., Sun, Y., Zhang, H., Shen, W., Cao, Y.: Towards high-resolution 3d anomalydetection:Ascalabledatasetandreal-timeframeworkforsubtleindustrial defects. In: Koenig, S., Jenkins, C., Taylor, M.E. (eds.) Fortieth AAAI Conference on Artificial Intelligence, Thirty-Eighth Conference on Innovative Applications of Artificial Intelligence, Sixteenth Sympo...

work page doi:10.1609/aaai.v40i5.37328 2026
[7]

In: IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2019, Long Beach, CA, USA, June 16-20, 2019

Choy, C.B., Gwak, J., Savarese, S.: 4d spatio-temporal convnets: Minkowski con- volutional neural networks. In: IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2019, Long Beach, CA, USA, June 16-20, 2019. pp. 3075–3084. Computer Vision Foundation / IEEE (2019).https://doi.org/10.1109/CVPR. 2019.00319, http://openaccess.thecvf.com/content_...

work page doi:10.1109/cvpr 2019
[8]

In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Deng, H., Li, X.: Anomaly detection via reverse distillation from one-class embed- ding. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022, New Orleans, LA, USA, June 18-24, 2022. pp. 9727–9736. IEEE (2022). https://doi.org/10.1109/CVPR52688.2022.00951 , https://doi.org/ 10.1109/CVPR52688.2022.00951

work page doi:10.1109/cvpr52688.2022.00951 2022
[9]

Frontiers of Engineering Management12(2), 343–360 (Jun 2025)

Du, J., Tao, C., Cao, X., Tsung, F.: 3d vision-based anomaly detection in man- ufacturing: A survey. Frontiers of Engineering Management12(2), 343–360 (Jun 2025). https://doi.org/10.1007/s42524- 025- 4189- 9 , https://doi.org/10. 1007/s42524-025-4189-9

work page doi:10.1007/s42524- 2025
[10]

In: IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2022, Waikoloa, HI, USA, January 3-8, 2022

Gudovskiy, D.A., Ishizaka, S., Kozuka, K.: CFLOW-AD: real-time unsuper- vised anomaly detection with localization via conditional normalizing flows. In: IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2022, Waikoloa, HI, USA, January 3-8, 2022. pp. 1819–1828. IEEE (2022).https://doi. org/10.1109/WACV51458.2022.00188 , https://doi.org/1...

work page doi:10.1109/wacv51458.2022.00188 2022
[11]

In: IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, CVPR 2023 - Workshops, Vancouver, BC, Canada, June 17-24, 2023

Horwitz, E., Hoshen, Y.: Back to the feature: Classical 3d features are (almost) all you need for 3d anomaly detection. In: IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, CVPR 2023 - Workshops, Vancouver, BC, Canada, June 17-24, 2023. pp. 2968–2977. IEEE (2023). https://doi.org/10.1109/ CVPRW59228.2023.00298, https://doi.org/10.1109/CVP...

work page doi:10.1109/cvprw59228.2023.00298 2023
[12]

IEEE Internet Things J

Huang, H., Wang, P., Pei, J., Wang, J., Alexanian, S., Niyato, D.: Deep learning ad- vancements in anomaly detection: A comprehensive survey. IEEE Internet Things J. 12(21), 44318–44342 (2025).https://doi.org/10.1109/JIOT.2025.3585884 , https://doi.org/10.1109/JIOT.2025.3585884

work page doi:10.1109/jiot.2025.3585884 2025
[13]

Springer Series in Statistics, Springer (1986)

Jolliffe, I.T.: Principal Component Analysis. Springer Series in Statistics, Springer (1986). https://doi.org/10.1007/978-1-4757-1904-8 , https://doi.org/10. 1007/978-1-4757-1904-8

work page doi:10.1007/978-1-4757-1904-8 1986
[14]

In: Proceedings of the 33rd ACM International Conference on Information and Knowledge Management

Kim, H., Lee, C.: Enhancing anomaly detection via generating diversified and hard- to-distinguish synthetic anomalies. In: Proceedings of the 33rd ACM International Conference on Information and Knowledge Management. p. 1089–1098. CIKM ’24, Association for Computing Machinery, New York, NY, USA (2024).https://doi. org/10.1145/3627673.3679623, https://doi....

work page doi:10.1145/3627673.3679623 2024
[15]

Bootstrapping SparseFormers from vision foundation models

Li, W., Xu, X., Gu, Y., Zheng, B., Gao, S., Wu, Y.: Towards scalable 3d anomaly detection and localization: A benchmark via 3d anomaly synthesis and A self- supervised learning network. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2024, Seattle, WA, USA, June 16-22, 2024. pp. 22207– 22216. IEEE (2024). https://doi.org/10.1109/C...

work page doi:10.1109/cvpr52733.2024.02096 2024
[16]

In: AAAI Conference on Artificial Intelligence (2024).https://doi.org/10.48550/arXiv.2412.13461

Liang, H., Xie, G., Hou, C., Wang, B., Gao, C., Wang, J.: Look inside for more: In- ternal spatial modality perception for 3d anomaly detection. In: AAAI Conference on Artificial Intelligence (2024).https://doi.org/10.48550/arXiv.2412.13461

work page doi:10.48550/arxiv.2412.13461 2024
[17]

Lin, Y., Chang, Y., Tong, X., Yu, J., Liotta, A., Huang, G., Song, W., Zeng, D., Wu, Z., Wang, Y., Zhang, W.: A survey on rgb, 3d, and multimodal approaches for unsupervised industrial image anomaly detection. Inf. Fusion121, 103139 (2025). https://doi.org/10.1016/J.INFFUS.2025.103139, https://doi.org/10.1016/ j.inffus.2025.103139

work page doi:10.1016/j.inffus.2025.103139 2025
[18]

In: Oh, A., Nau- mann, T., Globerson, A., Saenko, K., Hardt, M., Levine, S

Liu, J., Xie, G., Chen, R., Li, X., Wang, J., Liu, Y., Wang, C., Zheng, F.: Real3d-ad: A dataset of point cloud anomaly detection. In: Oh, A., Nau- mann, T., Globerson, A., Saenko, K., Hardt, M., Levine, S. (eds.) Advances in Neural Information Processing Systems 36: Annual Conference on Neu- ral Information Processing Systems 2023, NeurIPS 2023, New Orle...

2023
[19]

In: Avidan, S., Brostow, G.J., Cissé, M., Farinella, G.M., Hassner, T

Pang, Y., Wang, W., Tay, F.E.H., Liu, W., Tian, Y., Yuan, L.: Masked autoen- coders for point cloud self-supervised learning. In: Avidan, S., Brostow, G.J., Cissé, M., Farinella, G.M., Hassner, T. (eds.) Computer Vision - ECCV 2022 - 17th European Conference, Tel Aviv, Israel, October 23-27, 2022, Proceedings, Part II. Lecture Notes in Computer Science, v...

work page doi:10.1007/978-3-031-20086-1_35 2022
[20]

Expert Syst

Qin, J., Gu, C., Yu, J., Zhang, C.: Teacher-student network for 3d point cloud anomaly detection with few normal samples. Expert Syst. Appl. 228, 120371 (2023). https://doi.org/10.1016/J.ESWA.2023.120371 , https://doi.org/10. 1016/j.eswa.2023.120371 18 A. Balapour and F. Hach

work page doi:10.1016/j.eswa.2023.120371 2023
[21]

In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recog- nition (CVPR)

Roth, K., Pemula, L., Zepeda, J., Schölkopf, B., Brox, T., Gehler, P.V.: Towards to- tal recall in industrial anomaly detection. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022, New Orleans, LA, USA, June 18- 24, 2022. pp. 14298–14308. IEEE (2022).https://doi.org/10.1109/CVPR52688. 2022.01392, https://doi.org/10.1109/CVPR5268...

work page doi:10.1109/cvpr52688 2022
[22]

In: IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, CVPR 2023, Vancouver, BC, Canada, June 17-24,

Wang, Y., Peng, J., Zhang, J., Yi, R., Wang, Y., Wang, C.: Multimodal industrial anomaly detection via hybrid fusion. In: IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, CVPR 2023, Vancouver, BC, Canada, June 17-24,

2023
[23]

In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2023, Vancouver, BC, Canada, June 17-24, 2023

pp. 8032–8041. IEEE (2023).https://doi.org/10.1109/CVPR52729.2023. 00776, https://doi.org/10.1109/CVPR52729.2023.00776

work page doi:10.1109/cvpr52729.2023 2023
[24]

CoRRabs/2012.07122 (2020), https://arxiv.org/abs/ 2012.07122

Yang, J., Shi, Y., Qi, Z.: DFR: deep feature reconstruction for unsupervised anomaly segmentation. CoRRabs/2012.07122 (2020), https://arxiv.org/abs/ 2012.07122

work page arXiv 2012
[25]

In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2025, Nashville, TN, USA, June 11-15, 2025

Ye, J., Zhao, W., Yang, X., Cheng, G., Huang, K.: PO3AD: predicting point off- sets toward better 3d point cloud anomaly detection. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2025, Nashville, TN, USA, June 11-15, 2025. pp. 1353–1362. Computer Vision Foundation / IEEE (2025).https: / / doi . org / 10 . 1109 / CVPR52734 . 2025 ...

2025
[26]

In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Yu, X., Tang, L., Rao, Y., Huang, T., Zhou, J., Lu, J.: Point-bert: Pre-training 3d point cloud transformers with masked point modeling. In: IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, CVPR 2022, New Orleans, LA, USA, June 18-24, 2022. pp. 19291–19300. IEEE (2022).https://doi.org/ 10.1109/CVPR52688.2022.01871 , https://doi.org/10.110...

work page doi:10.1109/cvpr52688.2022.01871 2022
[27]

https://doi.org/10.48550/ARXIV.2510.16865 , https://doi.org/10

Yu,Y.,Chen,Z.,Xu,X.,Zhang, L.,Yang,H.,Nie,Y.,He,S.:Registrationisapow- erfulrotation-invariancelearnerfor3danomalydetection.CoRR abs/2510.16865 (2025). https://doi.org/10.48550/ARXIV.2510.16865 , https://doi.org/10. 48550/arXiv.2510.16865

work page doi:10.48550/arxiv.2510.16865 2025
[28]

The Bayesian approach to inverse problems

Zhou, Z., Wang, L., Fang, N., Wang, Z., Qiu, L., Zhang, S.: R3D-AD: recon- struction via diffusion for 3d anomaly detection. In: Leonardis, A., Ricci, E., Roth, S., Russakovsky, O., Sattler, T., Varol, G. (eds.) Computer Vision - ECCV 2024 - 18th European Conference, Milan, Italy, September 29-October 4, 2024, Proceedings, Part XXXVI. Lecture Notes in Com...

work page doi:10.1007/978- 2024
[29]

In: Cai, J., Kankanhalli, M.S., Prabhakaran, B., Boll, S., Subramanian, R., Zheng, L., Singh, V.K., César, P., Xie, L., Xu, D

Zhu, H., Xie, G., Hou, C., Dai, T., Gao, C., Wang, J., Shen, L.: Towards high- resolution 3d anomaly detection via group-level feature contrastive learning. In: Cai, J., Kankanhalli, M.S., Prabhakaran, B., Boll, S., Subramanian, R., Zheng, L., Singh, V.K., César, P., Xie, L., Xu, D. (eds.) Proceedings of the 32nd ACM International Conference on Multimedia...

work page doi:10.1145/3664647.3680919 2024
[30]

CoRR abs/2406.18941 (2024)

Zuo, Z., Dong, J., Wu, Y., Qu, Y., Wu, Z.: CLIP3D-AD: extending CLIP for 3d few-shot anomaly detection with multi-view images generation. CoRR abs/2406.18941 (2024). https : / / doi . org / 10 . 48550 / ARXIV . 2406 . 18941, https://doi.org/10.48550/arXiv.2406.18941 AF3AD 19 Supplementary Materials S1 Offset-Prediction Detector: Architectural Details This...

work page doi:10.48550/arxiv.2406.18941 2024

[1] [1]

CoRRabs/2112.09045 (2021), https://arxiv.org/abs/2112.09045

Bergmann, P., Jin, X., Sattlegger, D., Steger, C.: The mvtec 3d-ad dataset for un- supervised 3d anomaly detection and localization. CoRRabs/2112.09045 (2021), https://arxiv.org/abs/2112.09045

work page arXiv 2021

[2] [2]

Pattern Recognit

Cao, Y., Xu, X., Shen, W.: Complementary pseudo multimodal feature for point cloud anomaly detection. Pattern Recognit. 156, 110761 (2024). https://doi. org/10.1016/J.PATCOG.2024.110761 , https://doi.org/10.1016/j.patcog. 2024.110761

work page doi:10.1016/j.patcog.2024.110761 2024

[3] [3]

arXiv preprint arXiv:2401.16402 (2024)

Cao, Y., Xu, X., Zhang, J., Cheng, Y., Huang, X., Pang, G., Shen, W.: A survey on visual anomaly detection: Challenge, approach, and prospect. CoRR abs/2401.16402 (2024). https://doi.org/10.48550/ARXIV.2401.16402, https://doi.org/10.48550/arXiv.2401.16402

work page doi:10.48550/arxiv.2401.16402 2024

[4] [4]

ShapeNet: An Information-Rich 3D Model Repository

Chang, A.X., Funkhouser, T.A., Guibas, L.J., Hanrahan, P., Huang, Q., Li, Z., Savarese, S., Savva, M., Song, S., Su, H., Xiao, J., Yi, L., Yu, F.: Shapenet: An information-rich 3d model repository. CoRR abs/1512.03012 (2015), http:// arxiv.org/abs/1512.03012

work page internal anchor Pith review Pith/arXiv arXiv 2015

[5] [5]

IEEE Trans Autom

Cheng, Y., Cao, Y., Wang, D., Shen, W., Li, W.: Boosting global-local feature matching via anomaly synthesis for multi-class point cloud anomaly detection. IEEE Trans Autom. Sci. Eng. 22, 12560–12571 (2025).https://doi.org/10. 1109/TASE.2025.3544462, https://doi.org/10.1109/TASE.2025.3544462

work page doi:10.1109/tase.2025.3544462 2025

[6] [6]

In: Koenig, S., Jenkins, C., Taylor, M.E

Cheng, Y., Sun, Y., Zhang, H., Shen, W., Cao, Y.: Towards high-resolution 3d anomalydetection:Ascalabledatasetandreal-timeframeworkforsubtleindustrial defects. In: Koenig, S., Jenkins, C., Taylor, M.E. (eds.) Fortieth AAAI Conference on Artificial Intelligence, Thirty-Eighth Conference on Innovative Applications of Artificial Intelligence, Sixteenth Sympo...

work page doi:10.1609/aaai.v40i5.37328 2026

[7] [7]

In: IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2019, Long Beach, CA, USA, June 16-20, 2019

Choy, C.B., Gwak, J., Savarese, S.: 4d spatio-temporal convnets: Minkowski con- volutional neural networks. In: IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2019, Long Beach, CA, USA, June 16-20, 2019. pp. 3075–3084. Computer Vision Foundation / IEEE (2019).https://doi.org/10.1109/CVPR. 2019.00319, http://openaccess.thecvf.com/content_...

work page doi:10.1109/cvpr 2019

[8] [8]

In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Deng, H., Li, X.: Anomaly detection via reverse distillation from one-class embed- ding. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022, New Orleans, LA, USA, June 18-24, 2022. pp. 9727–9736. IEEE (2022). https://doi.org/10.1109/CVPR52688.2022.00951 , https://doi.org/ 10.1109/CVPR52688.2022.00951

work page doi:10.1109/cvpr52688.2022.00951 2022

[9] [9]

Frontiers of Engineering Management12(2), 343–360 (Jun 2025)

Du, J., Tao, C., Cao, X., Tsung, F.: 3d vision-based anomaly detection in man- ufacturing: A survey. Frontiers of Engineering Management12(2), 343–360 (Jun 2025). https://doi.org/10.1007/s42524- 025- 4189- 9 , https://doi.org/10. 1007/s42524-025-4189-9

work page doi:10.1007/s42524- 2025

[10] [10]

In: IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2022, Waikoloa, HI, USA, January 3-8, 2022

Gudovskiy, D.A., Ishizaka, S., Kozuka, K.: CFLOW-AD: real-time unsuper- vised anomaly detection with localization via conditional normalizing flows. In: IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2022, Waikoloa, HI, USA, January 3-8, 2022. pp. 1819–1828. IEEE (2022).https://doi. org/10.1109/WACV51458.2022.00188 , https://doi.org/1...

work page doi:10.1109/wacv51458.2022.00188 2022

[11] [11]

In: IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, CVPR 2023 - Workshops, Vancouver, BC, Canada, June 17-24, 2023

Horwitz, E., Hoshen, Y.: Back to the feature: Classical 3d features are (almost) all you need for 3d anomaly detection. In: IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, CVPR 2023 - Workshops, Vancouver, BC, Canada, June 17-24, 2023. pp. 2968–2977. IEEE (2023). https://doi.org/10.1109/ CVPRW59228.2023.00298, https://doi.org/10.1109/CVP...

work page doi:10.1109/cvprw59228.2023.00298 2023

[12] [12]

IEEE Internet Things J

Huang, H., Wang, P., Pei, J., Wang, J., Alexanian, S., Niyato, D.: Deep learning ad- vancements in anomaly detection: A comprehensive survey. IEEE Internet Things J. 12(21), 44318–44342 (2025).https://doi.org/10.1109/JIOT.2025.3585884 , https://doi.org/10.1109/JIOT.2025.3585884

work page doi:10.1109/jiot.2025.3585884 2025

[13] [13]

Springer Series in Statistics, Springer (1986)

Jolliffe, I.T.: Principal Component Analysis. Springer Series in Statistics, Springer (1986). https://doi.org/10.1007/978-1-4757-1904-8 , https://doi.org/10. 1007/978-1-4757-1904-8

work page doi:10.1007/978-1-4757-1904-8 1986

[14] [14]

In: Proceedings of the 33rd ACM International Conference on Information and Knowledge Management

Kim, H., Lee, C.: Enhancing anomaly detection via generating diversified and hard- to-distinguish synthetic anomalies. In: Proceedings of the 33rd ACM International Conference on Information and Knowledge Management. p. 1089–1098. CIKM ’24, Association for Computing Machinery, New York, NY, USA (2024).https://doi. org/10.1145/3627673.3679623, https://doi....

work page doi:10.1145/3627673.3679623 2024

[15] [15]

Bootstrapping SparseFormers from vision foundation models

Li, W., Xu, X., Gu, Y., Zheng, B., Gao, S., Wu, Y.: Towards scalable 3d anomaly detection and localization: A benchmark via 3d anomaly synthesis and A self- supervised learning network. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2024, Seattle, WA, USA, June 16-22, 2024. pp. 22207– 22216. IEEE (2024). https://doi.org/10.1109/C...

work page doi:10.1109/cvpr52733.2024.02096 2024

[16] [16]

In: AAAI Conference on Artificial Intelligence (2024).https://doi.org/10.48550/arXiv.2412.13461

Liang, H., Xie, G., Hou, C., Wang, B., Gao, C., Wang, J.: Look inside for more: In- ternal spatial modality perception for 3d anomaly detection. In: AAAI Conference on Artificial Intelligence (2024).https://doi.org/10.48550/arXiv.2412.13461

work page doi:10.48550/arxiv.2412.13461 2024

[17] [17]

Lin, Y., Chang, Y., Tong, X., Yu, J., Liotta, A., Huang, G., Song, W., Zeng, D., Wu, Z., Wang, Y., Zhang, W.: A survey on rgb, 3d, and multimodal approaches for unsupervised industrial image anomaly detection. Inf. Fusion121, 103139 (2025). https://doi.org/10.1016/J.INFFUS.2025.103139, https://doi.org/10.1016/ j.inffus.2025.103139

work page doi:10.1016/j.inffus.2025.103139 2025

[18] [18]

In: Oh, A., Nau- mann, T., Globerson, A., Saenko, K., Hardt, M., Levine, S

Liu, J., Xie, G., Chen, R., Li, X., Wang, J., Liu, Y., Wang, C., Zheng, F.: Real3d-ad: A dataset of point cloud anomaly detection. In: Oh, A., Nau- mann, T., Globerson, A., Saenko, K., Hardt, M., Levine, S. (eds.) Advances in Neural Information Processing Systems 36: Annual Conference on Neu- ral Information Processing Systems 2023, NeurIPS 2023, New Orle...

2023

[19] [19]

In: Avidan, S., Brostow, G.J., Cissé, M., Farinella, G.M., Hassner, T

Pang, Y., Wang, W., Tay, F.E.H., Liu, W., Tian, Y., Yuan, L.: Masked autoen- coders for point cloud self-supervised learning. In: Avidan, S., Brostow, G.J., Cissé, M., Farinella, G.M., Hassner, T. (eds.) Computer Vision - ECCV 2022 - 17th European Conference, Tel Aviv, Israel, October 23-27, 2022, Proceedings, Part II. Lecture Notes in Computer Science, v...

work page doi:10.1007/978-3-031-20086-1_35 2022

[20] [20]

Expert Syst

Qin, J., Gu, C., Yu, J., Zhang, C.: Teacher-student network for 3d point cloud anomaly detection with few normal samples. Expert Syst. Appl. 228, 120371 (2023). https://doi.org/10.1016/J.ESWA.2023.120371 , https://doi.org/10. 1016/j.eswa.2023.120371 18 A. Balapour and F. Hach

work page doi:10.1016/j.eswa.2023.120371 2023

[21] [21]

In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recog- nition (CVPR)

Roth, K., Pemula, L., Zepeda, J., Schölkopf, B., Brox, T., Gehler, P.V.: Towards to- tal recall in industrial anomaly detection. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022, New Orleans, LA, USA, June 18- 24, 2022. pp. 14298–14308. IEEE (2022).https://doi.org/10.1109/CVPR52688. 2022.01392, https://doi.org/10.1109/CVPR5268...

work page doi:10.1109/cvpr52688 2022

[22] [22]

In: IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, CVPR 2023, Vancouver, BC, Canada, June 17-24,

Wang, Y., Peng, J., Zhang, J., Yi, R., Wang, Y., Wang, C.: Multimodal industrial anomaly detection via hybrid fusion. In: IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, CVPR 2023, Vancouver, BC, Canada, June 17-24,

2023

[23] [23]

In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2023, Vancouver, BC, Canada, June 17-24, 2023

pp. 8032–8041. IEEE (2023).https://doi.org/10.1109/CVPR52729.2023. 00776, https://doi.org/10.1109/CVPR52729.2023.00776

work page doi:10.1109/cvpr52729.2023 2023

[24] [24]

CoRRabs/2012.07122 (2020), https://arxiv.org/abs/ 2012.07122

Yang, J., Shi, Y., Qi, Z.: DFR: deep feature reconstruction for unsupervised anomaly segmentation. CoRRabs/2012.07122 (2020), https://arxiv.org/abs/ 2012.07122

work page arXiv 2012

[25] [25]

In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2025, Nashville, TN, USA, June 11-15, 2025

Ye, J., Zhao, W., Yang, X., Cheng, G., Huang, K.: PO3AD: predicting point off- sets toward better 3d point cloud anomaly detection. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2025, Nashville, TN, USA, June 11-15, 2025. pp. 1353–1362. Computer Vision Foundation / IEEE (2025).https: / / doi . org / 10 . 1109 / CVPR52734 . 2025 ...

2025

[26] [26]

In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Yu, X., Tang, L., Rao, Y., Huang, T., Zhou, J., Lu, J.: Point-bert: Pre-training 3d point cloud transformers with masked point modeling. In: IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, CVPR 2022, New Orleans, LA, USA, June 18-24, 2022. pp. 19291–19300. IEEE (2022).https://doi.org/ 10.1109/CVPR52688.2022.01871 , https://doi.org/10.110...

work page doi:10.1109/cvpr52688.2022.01871 2022

[27] [27]

https://doi.org/10.48550/ARXIV.2510.16865 , https://doi.org/10

Yu,Y.,Chen,Z.,Xu,X.,Zhang, L.,Yang,H.,Nie,Y.,He,S.:Registrationisapow- erfulrotation-invariancelearnerfor3danomalydetection.CoRR abs/2510.16865 (2025). https://doi.org/10.48550/ARXIV.2510.16865 , https://doi.org/10. 48550/arXiv.2510.16865

work page doi:10.48550/arxiv.2510.16865 2025

[28] [28]

The Bayesian approach to inverse problems

Zhou, Z., Wang, L., Fang, N., Wang, Z., Qiu, L., Zhang, S.: R3D-AD: recon- struction via diffusion for 3d anomaly detection. In: Leonardis, A., Ricci, E., Roth, S., Russakovsky, O., Sattler, T., Varol, G. (eds.) Computer Vision - ECCV 2024 - 18th European Conference, Milan, Italy, September 29-October 4, 2024, Proceedings, Part XXXVI. Lecture Notes in Com...

work page doi:10.1007/978- 2024

[29] [29]

In: Cai, J., Kankanhalli, M.S., Prabhakaran, B., Boll, S., Subramanian, R., Zheng, L., Singh, V.K., César, P., Xie, L., Xu, D

Zhu, H., Xie, G., Hou, C., Dai, T., Gao, C., Wang, J., Shen, L.: Towards high- resolution 3d anomaly detection via group-level feature contrastive learning. In: Cai, J., Kankanhalli, M.S., Prabhakaran, B., Boll, S., Subramanian, R., Zheng, L., Singh, V.K., César, P., Xie, L., Xu, D. (eds.) Proceedings of the 32nd ACM International Conference on Multimedia...

work page doi:10.1145/3664647.3680919 2024

[30] [30]

CoRR abs/2406.18941 (2024)

Zuo, Z., Dong, J., Wu, Y., Qu, Y., Wu, Z.: CLIP3D-AD: extending CLIP for 3d few-shot anomaly detection with multi-view images generation. CoRR abs/2406.18941 (2024). https : / / doi . org / 10 . 48550 / ARXIV . 2406 . 18941, https://doi.org/10.48550/arXiv.2406.18941 AF3AD 19 Supplementary Materials S1 Offset-Prediction Detector: Architectural Details This...

work page doi:10.48550/arxiv.2406.18941 2024