BridgeNet: A Unified Multimodal Framework for Bridging 2D and 3D Industrial Anomaly Detection

An Xiang; Cheng-zhong Xu; Kejiang Ye; Xitong Gao; Zixuan Huang

REVIEW 3 major objections 5 minor 46 references

BridgeNet: A Unified Multimodal Framework for Bridging 2D and 3D Industrial Anomaly Detection

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read BridgeNet claims one shared 2D network, fed RGB and depth images, can beat dedicated 3D anomaly-detection pipelines.

desk verdict A simple, effective parameter-shared depth-image recipe for RGB+3D anomaly detection, held back mainly by test-set hyperparameter tuning and a missing 3D-feature ablation. read the letter →

arxiv 2507.19253 v1 pith:7WISMM26 submitted 2025-07-25 cs.CV

An Xiang , Zixuan Huang , Xitong Gao , Kejiang Ye , Cheng-zhong Xu This is my paper · ORCID

classification cs.CV

keywords multimodalindustrialanomalydetectiondepthimagerepresentationparametersharinggenerationmulti-scaleGaussiannoisetexturesynthesisMVTec-3DADfew-shot

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

BridgeNet claims that 3D industrial anomaly detection can be handled by a single 2D network if the point cloud is first turned into a depth image and that depth image is processed by the same ImageNet-pretrained backbone as the RGB image. The paper argues that sharing all parameters implicitly aligns the two modalities, so a simple feature concatenation plus a linear fusion layer replaces explicit cross-modal alignment modules. To supply the scarce anomaly examples, it generates synthetic anomalies in both modalities: multi-scale Gaussian noise injected at three model depths with decreasing variance, and DTD texture patches pasted onto both RGB and depth images. On this recipe the method reports state-of-the-art detection on MVTec-3D AD (image-level AUROC 99.3%, P-AUPRO 97.7%) and on Eyecandies, plus strong few-shot performance with as few as five training samples per class. If true, the practical consequence is that 3D anomaly detection does not require point-cloud backbones or complex fusion.

What carries the argument

The load-bearing object is the depth image obtained by dropping the x and y components of the ordered point cloud and keeping only the z values, after filling missing depth values and separating foreground from background by distance to a plane. This preprocessing step turns 3D geometry into an image that a standard 2D backbone can read, and it makes depth holes and deformations appear as visible texture-like patterns. On top of it, the framework uses a parameter-shared WideResNet-50 pretrained on ImageNet as the feature extractor for both modalities, a bias-free fully connected fusion adaptor, the Multi-Scale Gaussian Anomaly Generator (MGAG) that injects Gaussian noise of decreasing variance at three depths of the model, the Unified Texture Anomaly Generator (UTAG) that pastes DTD textures onto RGB and depth foregrounds with an opacity parameter, and a dual-modal discriminator trained with BCE loss for normal and Gaussian-anomaly masks plus focal loss for texture-anomaly masks. The machinery works by letting the shared backbone do the modality alignment implicitly, so no separate alignment network is needed.

What would settle it

Take a set of MVTec-3D AD anomalies and remove those that are visible in the full point cloud but not in the rendered z-depth image, for instance shallow indentations smaller than the depth-image resolution or defects that only change local surface orientation. If BridgeNet's detection on that filtered set drops to the level of an RGB-only baseline, the depth-image representation is the bottleneck; a synthetic version would render a point cloud with a purely curvature-based anomaly and check whether the depth image changes at all.

Watch

Extended reading notes

Core claim

The central claim is that the gap between 2D and 3D anomaly detection is mostly a representation problem, and that converting ordered point clouds into three-channel depth images makes the two modalities similar enough for one shared network. Using a shared ImageNet-pretrained feature extractor on RGB and depth, concatenating the multi-scale features, and passing them through a single linear fusion adaptor produces feature distributions that are close across modalities, as the paper shows by histograms of per-dimension standard deviations. The paper then claims two generators create useful training anomalies: MGAG adds zero-mean Gaussian noise at the image level, at the pre-fusion feature level, and at the post-fusion adapted-feature level, with variances decreasing from shallow to deep (best configuration 0.12, 0.04, 0.02), because deeper feature spaces are more compact; and UTAG pastes textures from the DTD dataset onto masked foreground regions of both RGB and grayscale-converted depth images. A selective-modality rule randomly makes each synthesized anomaly RGB-only, depth-only, or both, preventing the model from over-relying on one modality. The authors report that this framework, with all parameters shared, reaches 99.3% I-AUROC and 97.7% P-AUPRO on MVTec-3D AD, outperforming prior 3D and RGB+3D methods, and that the same design transfers to Eyecandies and to 2D-only datasets.

Load-bearing premise

The method depends on the idea that a z-only depth image preserves the geometric information needed to catch depth-only anomalies; if holes, dents, or deformations disappear when the point cloud is rendered into a depth image, the shared 2D pipeline loses its 3D advantage.

Editorial extensions

If this is right

Point-cloud backbones and explicit cross-modal alignment become optional: under the reported setup the framework runs at 25 FPS using 2289 MB of memory, so 3D industrial anomaly detection inherits the efficiency of 2D methods.
With 50 samples per class, BridgeNet reaches 97.7% I-AUROC and 97.2% P-AUPRO on MVTec-3D AD, so the same anomaly generators support few-shot deployment.
The MGAG module also improves 2D-only anomaly detection, lifting SimpleNet and GLASS on MVTec AD and VisA, which suggests scale-decreasing multi-layer Gaussian noise is a generally useful training signal.
Adopting the depth-image preprocessing and shared-parameter scheme improves existing multimodal methods such as AST+ and CFM+ even when their fusion modules are unchanged.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The paper does not test anomalies that are visible in the full point cloud but invisible in a z-only depth rendering, such as dents smaller than the depth resolution or defects defined purely by surface curvature; a benchmark containing such cases would clarify how much 3D information survives the depth-image representation.
The selective-modality rule is effectively a modality-dropout regularizer, and the same idea could transfer to other multimodal tasks where one channel is sometimes uninformative or missing.
The decreasing-noise finding suggests feature-space volume shrinks as information passes through the fusion adaptor; this could be measured directly and used to set noise scales in other dense-prediction tasks.
If the depth-image representation is sufficient, 3D anomaly detection may converge with the 2D literature faster than expected, because new 2D generators and backbones could be reused without redesign.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Read through BridgeNet. It's a clean, parameter-shared framework: point clouds become depth images, one ImageNet-pretrained 2D backbone extracts features for both RGB and depth, and anomaly generation is unified across modalities via multi-scale Gaussian noise at three model levels, DTD textures on both sides, and a random 'selective modality' switch. The composition is new; the ingredients are not, but the combination is sensible and the engineering is solid. Results are strong: SOTA on MVTec-3D AD in 3D, RGB, and combined settings, SOTA I-AUROC on Eyecandies, with better memory/FPS than most point-cloud-based methods. The ablations are reasonably thorough, including AST+/CFM+ comparisons that hold the depth-image preprocessing fixed and a UTAG-vs-Perlin comparison.

The soft spots are more about empirical hygiene than the core idea. The final noise scales, injection levels, and α are picked from ablations on the same test set (Tables 5–7), so the headline numbers are somewhat optimistic. The few-shot results are single-run with no error bars; the 5-shot and 10-shot edges could reflect sampling luck. And the stress-test concern lands: the paper never compares its depth-image 3D stream against an explicit 3D feature pipeline (FPFH, PointMAE) with everything else held fixed. Table 8 shows that sharing a 2D backbone beats prior backbones when you already use depth images, but it doesn't tell you whether the depth-image representation preserves enough geometry. The 3D-only SOTA claim rests on a modest 1.3% mean gap, and on several categories BridgeNet is not the best.

The UTAG contribution is better supported: Table 10 shows clear gains over Perlin noise in both EasyNet and BridgeNet. The MGAG multi-scale claim is weaker—adding G1 only yields 0.1–0.3% over G2+G3—but the paper does not oversell it.

Overall, the central result (simple shared-parameter depth-image method beats prior point-cloud fusion approaches) holds up. I'd want the authors to run the ablation against explicit 3D features, report few-shot variance, and confirm hyperparameters on a validation split, but those are standard referee requests, not fundamental flaws. The paper deserves a serious referee.

Referee Report

3 major / 5 minor

Summary. BridgeNet proposes a unified multimodal framework for 2D/3D industrial anomaly detection. It converts aligned point clouds into three-channel depth images, feeds RGB and depth through a shared ImageNet-pretrained ResNet, concatenates multi-scale features, and passes them through a fusion adaptor. During training, it synthesizes anomalies with a multi-scale Gaussian anomaly generator (noise at image, feature, and adapted-feature levels) and a DTD-based texture anomaly generator applied to both modalities, with selective single/both-modality injection. A shared dual-modal discriminator is trained with BCE and focal losses. On MVTec-3D AD the authors report I-AUROC 99.3% and P-AUPRO 97.7% in the combined setting, SOTA in the 3D, RGB, and combined columns, and also report SOTA on Eyecandies and few-shot results, with code released.

Significance. The paper addresses an important problem and demonstrates a practical, unified recipe: the depth-image representation avoids point-cloud backbones and gives large efficiency gains (Table 11), and the ablation in Table 8 shows that sharing a 2D backbone improves AST and CFM when the preprocessing is fixed. Strengths include the breadth of comparison (two benchmarks, few-shot, 2D generalization in Appendix D), release of code, and the directness of the synthetic anomaly generators. If the reported numbers hold under proper evaluation protocols, the framework is a valuable contribution. The main reservations concern evaluation rigor: hyperparameters are tuned on the test set, results are single-run, and the central depth-image representation is never compared with explicit 3D features under matched conditions.

major comments (3)

[§4.2, Tables 6–7] The final configuration (noise scales σ1=0.12, σ2=0.04, σ3=0.02; injection at G1, G2, G3; α=1/3) is selected by comparing I-AUROC/P-AUPRO values on the MVTec-3D AD test set. Since the reported headline results are the best configuration from this in-test-set search, the claimed SOTA margins (e.g., +1.5% over 3DSR in Table 1) are optimistically biased. The dataset provides a 294-sample validation split; the authors should tune on it, or at minimum report the full ablation grid with a clear model-selection protocol and confidence intervals, so the reader can see how much of the gain is configuration search.
[§4.1, Tables 1–4; §B, Tables 13–14] All reported metrics appear to come from a single training run, and the few-shot experiments do not report variance over multiple random training-sample draws. Several per-category differences and some mean differences between BridgeNet and the next-best method are small (1–2 points), so without seeds and standard deviations it is impossible to tell whether the SOTA claim is statistically meaningful. The authors should report mean±std over at least 3–5 seeds for the main tables, and for the few-shot table they should resample the training set multiple times and report the distribution of metrics.
[§3.1, Tables 1 and 8] The design premise that a depth image processed by a shared 2D backbone preserves the 3D geometric cues needed for depth-only anomalies is asserted but never directly tested. Table 8 varies the downstream method while keeping the depth-image preprocessing fixed, so it cannot isolate the contribution of the depth-image representation; comparison methods such as M3DM and LSFA, which use point-cloud backbones, differ in many other respects. To make the 3D-only results (Table 1: 0.935 I-AUROC, 0.945 P-AUPRO) attributable to the proposed representation, the authors should add an ablation that feeds explicit 3D features (e.g., FPFH or a point-cloud branch) into the same MGAG/UTAG/discriminator pipeline, or otherwise demonstrate that no critical geometric information is lost in the depth-image conversion.

minor comments (5)

[§3.4, Eq. (10)] The text says 'Sigmod' where it should say 'Sigmoid'.
[§3.2–3.3, Eqs. (6) and (9)] The random variable p used to select modality combinations is never defined; state that p is sampled uniformly from (0,1) and that α=1/3 yields equal-probability branches.
[§C, Tables 15–16] On Eyecandies, the 3D and P-AUPRO results are below LDM; please qualify the SOTA claim to specify the combined-setting I-AUROC and describe the localization results as competitive rather than SOTA.
[Appendix B] The statement that 50-shot P-AUPRO 'matches' SOTA is inconsistent with Table 3, where BridgeNet's 0.972 exceeds LSFA's 0.962; align the text with the table.
[Figure 6] The histograms of feature standard deviations do not directly demonstrate that RGB and depth distributions are 'similar and close'; a quantitative distributional distance (e.g., MMD or Wasserstein distance) would support the implicit-alignment claim more convincingly.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: BridgeNet's SOTA claims are supported by external benchmark evaluation rather than derived from its own assumptions.

full rationale

BridgeNet's central claims—SOTA I-AUROC and P-AUPRO on MVTec-3D AD and Eyecandies—are empirical results computed on held-out test sets of external benchmarks and compared against published methods. The architectural choices (depth-image projection, shared-parameter backbone, MGAG, UTAG) are inputs to the experiments, not outputs derived from the benchmark numbers. No equation defines a predicted quantity in terms of a fitted quantity: the only free hyperparameters (alpha, beta, noise scales) are fixed to stated values and their effects are ablated in Tables 5–7. The paper contains no self-citations by the authors and invokes no uniqueness theorem from prior work. Appendix B honestly reports a 1-shot weakness, which is a limitation, not circularity. The closest concern—synthetic anomaly generators being tuned on the evaluation benchmark—is hyperparameter/model selection and does not make the benchmark results equivalent to the method's inputs by construction. Therefore no circular step is present.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

No free parameters beyond the standard network hyperparameters; the listed values are the ones directly selected on the benchmark test sets. No new entities are introduced.

free parameters (4)

Gaussian noise scales sigma1, sigma2, sigma3 = 0.12, 0.04, 0.02
Chosen via ablation on MVTec-3D AD (Table 7); the paper states that adding larger noise at shallower levels improves scores.
Alpha for Selective Modality = 1/3
Set to 1/3 without derivation; controls probability of per-modality anomaly injection.
Texture opacity beta distribution = N(0.5, 0.3), truncated to (0.2, 0.8)
Borrowed from prior work [24] according to the paper; controls blending strength of texture anomalies.
Choice of Gaussian injection levels (G1, G2, G3) = all three levels, with sigma1 > sigma2 > sigma3
Ablation Table 6 selects the configuration; this is model selection on the evaluation benchmark.

assumptions (4)

domain assumption Point-cloud-to-depth-image conversion preserves the 3D anomaly signal; depth anomalies appear as visible texture in the rendered depth image.
Section 3.1 and the UTAG description (Figure 4) rely on this to use 2D backbones on depth.
domain assumption ImageNet-pretrained features are informative for depth images.
Section 3.1 uses a ResNet-like ImageNet pretrained model for both RGB and depth.
ad hoc to paper Sharing parameters between RGB and depth implicitly aligns the two modalities.
Section 3.1 claims this and shows histograms (Figure 2), but no formal or thorough empirical proof is given.
ad hoc to paper Synthetic Gaussian and texture anomalies at the chosen scales sufficiently approximate real industrial anomalies.
Section 3.2 and 3.3 design anomaly generators based on this assumption; the paper acknowledges the gap to real anomalies for Perlin-based methods and claims its own generators are closer, but only via benchmark results.

how reviews work

0 comments

Cite this review

Pith. "Pith review of BridgeNet: A Unified Multimodal Framework for Bridging 2D and 3D Industrial Anomaly Detection." pith.science (2026). https://pith.science/paper/7WISMM26

@misc{pith2026250719253,
  author       = {Pith},
  title        = {Pith review of: BridgeNet: A Unified Multimodal Framework for Bridging 2D and 3D Industrial Anomaly Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7WISMM26}},
  note         = {Machine review of arXiv:2507.19253}
}

read the original abstract

Industrial anomaly detection for 2D objects has gained significant attention and achieved progress in anomaly detection (AD) methods. However, identifying 3D depth anomalies using only 2D information is insufficient. Despite explicitly fusing depth information into RGB images or using point cloud backbone networks to extract depth features, both approaches struggle to adequately represent 3D information in multimodal scenarios due to the disparities among different modal information. Additionally, due to the scarcity of abnormal samples in industrial data, especially in multimodal scenarios, it is necessary to perform anomaly generation to simulate real-world abnormal samples. Therefore, we propose a novel unified multimodal anomaly detection framework to address these issues. Our contributions consist of 3 key aspects. (1) We extract visible depth information from 3D point cloud data simply and use 2D RGB images to represent appearance, which disentangles depth and appearance to support unified anomaly generation. (2) Benefiting from the flexible input representation, the proposed Multi-Scale Gaussian Anomaly Generator and Unified Texture Anomaly Generator can generate richer anomalies in RGB and depth. (3) All modules share parameters for both RGB and depth data, effectively bridging 2D and 3D anomaly detection. Subsequent modules can directly leverage features from both modalities without complex fusion. Experiments show our method outperforms state-of-the-art (SOTA) on MVTec-3D AD and Eyecandies datasets. Code available at: https://github.com/Xantastic/BridgeNet

Figures

Figures reproduced from arXiv: 2507.19253 by the authors.

**Figure 3.** Overview of the proposed BridgeNet. In the training phase, RGB and depth images are inputted into shared [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

**Figure 4.** Diagram of the Unified Texture Anomaly Generator. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗

Figures from the paper (4 more)

**Figure 5.** Figure 5: Qualitative results of BridgeNet on the MVTec-3D [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

**Figure 6.** Figure 6: Histograms of standard deviations per dimension [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

**Figure 7.** Figure 7: Qualitative results of BridgeNet on the MVTec-3D [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]

**Figure 8.** Figure 8: Qualitative results of BridgeNet on the Eyecandies dataset. [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 25 canonical work pages

[1]

Kilian Batzner, Lars Heckler, and Rebecca König. 2024. Efficientad: Accurate visual anomaly detection at millisecond-level latencies. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision . 128–138

work page 2024
[2]

Paul Bergmann, Kilian Batzner, Michael Fauser, David Sattlegger, and Carsten Steger. 2022. Beyond dents and scratches: Logical constraints in unsupervised anomaly detection and localization. International Journal of Computer Vision 130, 4 (2022), 947–969

2022
[3]

Paul Bergmann, Michael Fauser, David Sattlegger, and Carsten Steger. 2019. MVTec AD–A comprehensive real-world dataset for unsupervised anomaly de- tection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 9592–9600

2019
[4]

Paul Bergmann, Michael Fauser, David Sattlegger, and Carsten Steger. 2020. Uninformed students: Student-teacher anomaly detection with discriminative latent embeddings. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 4183–4192

2020
[5]

Paul Bergmann, Xin Jin, David Sattlegger, and Carsten Steger. 2021. The mvtec 3d-ad dataset for unsupervised 3d anomaly detection and localization. arXiv preprint arXiv:2112.09045 (2021)

arXiv 2021
[6]

Luca Bonfiglioli, Marco Toschi, Davide Silvestri, Nicola Fioraio, and Daniele De Gregorio. 2022. The eyecandies dataset for unsupervised multimodal anomaly detection and localization. In Proceedings of the Asian Conference on Computer Vision. 3586–3602

work page 2022
[7]

Yunkang Cao, Xiaohao Xu, and Weiming Shen. 2024. Complementary pseudo multimodal feature for point cloud anomaly detection. Pattern Recognition 156 (2024), 110761

work page 2024
[8]

Qiyu Chen, Huiyuan Luo, Chengkan Lv, and Zhengtao Zhang. 2024. A unified anomaly synthesis strategy with gradient ascent for industrial anomaly detection and localization. arXiv preprint arXiv:2407.09359 (2024)

arXiv 2024

Show all 46 references

[9]

Ruitao Chen, Guoyang Xie, Jiaqi Liu, Jinbao Wang, Ziqi Luo, Jinfan Wang, and Feng Zheng. 2023. Easynet: An easy network for 3d industrial anomaly detection. In Proceedings of the 31st ACM International Conference on Multimedia. 7038–7046

2023
[10]

Yu-Min Chu, Chieh Liu, Ting-I Hsieh, Hwann-Tzong Chen, and Tyng-Luh Liu
[11]

Mircea Cimpoi, Subhransu Maji, Iasonas Kokkinos, Sammy Mohamed, and An- drea Vedaldi. 2014. Describing textures in the wild. In Proceedings of the IEEE conference on computer vision and pattern recognition . 3606–3613

2014
[12]

Niv Cohen and Yedid Hoshen. 2020. Sub-image anomaly detection with deep pyramid correspondences. arXiv preprint arXiv:2005.02357 (2020)

2020 arXiv
[13]

Alex Costanzino, Pierluigi Zama Ramirez, Giuseppe Lisanti, and Luigi Di Stefano
[14]

Thomas Defard, Aleksandr Setkov, Angelique Loesch, and Romaric Audigier
[15]

Hanqiu Deng and Xingyu Li. 2022. Anomaly detection via reverse distillation from one-class embedding. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 9737–9746

2022
[16]

Dong Gong, Lingqiao Liu, Vuong Le, Budhaditya Saha, Moussa Reda Mansour, Svetha Venkatesh, and Anton van den Hengel. 2019. Memorizing normality to detect anomaly: Memory-augmented deep autoencoder for unsupervised anomaly detection. In Proceedings of the IEEE/CVF international...

2019
[17]

Zhihao Gu, Liang Liu, Xu Chen, Ran Yi, Jiangning Zhang, Yabiao Wang, Chengjie Wang, Annan Shu, Guannan Jiang, and Lizhuang Ma. 2023. Remembering normal- ity: Memory-guided knowledge distillation for unsupervised anomaly detection. In Proceedings of the IEEE/CVF International C...

2023
[18]

Eliahu Horwitz and Yedid Hoshen. 2023. Back to the feature: classical 3d features are (almost) all you need for 3d anomaly detection. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 2968–2977

2023
[19]

Jiarui Lei, Xiaobo Hu, Yue Wang, and Dong Liu. 2023. Pyramidflow: High- resolution defect contrastive localization using pyramid normalizing flow. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 14143–14152

2023
[20]

Chun-Liang Li, Kihyuk Sohn, Jinsung Yoon, and Tomas Pfister. 2021. Cutpaste: Self-supervised learning for anomaly detection and localization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 9664–9674

2021
[21]

Kecen Li, Bingquan Dai, Jingjing Fu, and Xinwen Hou. 2024. DAS3D: Dual- modality Anomaly Synthesis for 3D Anomaly Detection. In arXiv preprint arXiv:2410.09821

2024 arXiv
[22]

Chieh Liu, Yu-Min Chu, Ting-I Hsieh, Hwann-Tzong Chen, and Tyng-Luh Liu
[23]

Wenrui Liu, Hong Chang, Bingpeng Ma, Shiguang Shan, and Xilin Chen. 2023. Diversity-measurable anomaly detection. In Proceedings of the IEEE/CVF confer- ence on computer vision and pattern recognition . 12147–12156

2023
[24]

Zhikang Liu, Yiming Zhou, Yuansheng Xu, and Zilei Wang. 2023. Simplenet: A simple network for image anomaly detection and localization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 20402– 20411

2023
[25]

Pramuditha Perera, Ramesh Nallapati, and Bing Xiang. 2019. Ocgan: One-class novelty detection using gans with constrained latent representations. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition . 2898–2906

2019
[26]

T-YLPG Ross and GKHP Dollár. 2017. Focal loss for dense object detection. In proceedings of the IEEE conference on computer vision and pattern recognition . 2980–2988

2017
[27]

Karsten Roth, Latha Pemula, Joaquin Zepeda, Bernhard Schölkopf, Thomas Brox, and Peter Gehler. 2022. Towards total recall in industrial anomaly detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 14318–14328

2022
[28]

Marco Rudolph, Tom Wehrbein, Bodo Rosenhahn, and Bastian Wandt. 2022. Fully convolutional cross-scale-flows for image-based defect detection. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. 1088–1097. MM ’25, October 27–31, 2025, Dublin, Ir...

2022
[29]

Marco Rudolph, Tom Wehrbein, Bodo Rosenhahn, and Bastian Wandt. 2023. Asymmetric student-teacher networks for industrial anomaly detection. In Pro- ceedings of the IEEE/CVF winter conference on applications of computer vision . 2592–2602

2023
[30]

Joan Serrà, David Álvarez, Vicenç Gómez, Olga Slizovskaia, José F Núñez, and Jordi Luque. 2019. Input complexity and out-of-distribution detection with likelihood-based generative models. arXiv preprint arXiv:1909.11480 (2019)

2019 arXiv
[31]

Tran Dinh Tien, Anh Tuan Nguyen, Nguyen Hoang Tran, Ta Duc Huy, Soan Duong, Chanh D Tr Nguyen, and Steven QH Truong. 2023. Revisiting reverse distillation for anomaly detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 24511–24520

2023
[32]

Yuanpeng Tu, Boshen Zhang, Liang Liu, Yuxi Li, Jiangning Zhang, Yabiao Wang, Chengjie Wang, and Cairong Zhao. 2024. Self-supervised feature adaptation for 3d industrial anomaly detection. In European Conference on Computer Vision . Springer, 75–91

2024
[33]

Yue Wang, Jinlong Peng, Jiangning Zhang, Ran Yi, Yabiao Wang, and Chengjie Wang. 2023. Multimodal industrial anomaly detection via hybrid fusion. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 8032–8041

2023
[34]

Qiangwei Wu, Hui Li, Chenyu Tian, Long Wen, and Xinyu Li. 2024. AEKD: Un- supervised auto-encoder knowledge distillation for industrial anomaly detection. Journal of Manufacturing Systems 73 (2024), 159–169

2024
[35]

Vitjan Zavrtanik, Matej Kristan, and Danijel Skočaj. 2021. Draem-a discrimi- natively trained reconstruction embedding for surface anomaly detection. In Proceedings of the IEEE/CVF international conference on computer vision . 8330– 8339

2021
[36]

Vitjan Zavrtanik, Matej Kristan, and Danijel Skočaj. 2021. Reconstruction by inpainting for visual anomaly detection. Pattern Recognition 112 (2021), 107706

2021
[37]

Vitjan Zavrtanik, Matej Kristan, and Danijel Skočaj. 2022. Dsr–a dual subspace re-projection network for surface anomaly detection. In European conference on computer vision. Springer, 539–554

2022
[38]

Vitjan Zavrtanik, Matej Kristan, and Danijel Skočaj. 2024. Cheating depth: Enhancing 3d surface anomaly detection via depth simulation. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision . 2164–2172

2024
[39]

Vitjan Zavrtanik, Matej Kristan, and Danijel Skočaj. 2024. Keep dræming: discrim- inative 3d anomaly detection through anomaly simulation. Pattern Recognition Letters 181 (2024), 113–119

2024
[40]

Yixuan Zhou, Xing Xu, Jingkuan Song, Fumin Shen, and Heng Tao Shen. 2024. Msflow: Multiscale flow-based framework for unsupervised anomaly detection. IEEE Transactions on Neural Networks and Learning Systems (2024)

2024
[41]

Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang, and Onkar Dabeer
[2021]

In International Conference on Pattern Recognition

Padim: a patch distribution modeling framework for anomaly detection and localization. In International Conference on Pattern Recognition . Springer, 475–489
[2022]

In European Conference on Computer Vision

Spot-the-difference self-supervised pre-training for anomaly detection and segmentation. In European Conference on Computer Vision . Springer, 392–408. BridgeNet: A Unified Multimodal Framework for Bridging 2D and 3D Industrial Anomaly Detection MM ’25, October 27–31, 2025, Du...

2025
[2023]

In Proceed- ings of the 40th International Conference on Machine Learning

Shape-guided dual-memory learning for 3D anomaly detection. In Proceed- ings of the 40th International Conference on Machine Learning . 6185–6194
[2024]

In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

Multimodal industrial anomaly detection by crossmodal feature mapping. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 17234–17243
[2025]

In European Conference on Computer Vision

Learning Diffusion Models for Multi-view Anomaly Detection. In European Conference on Computer Vision . Springer, 328–345

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

Kilian Batzner, Lars Heckler, and Rebecca König. 2024. Efficientad: Accurate visual anomaly detection at millisecond-level latencies. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision . 128–138

work page 2024

[2] [2]

Paul Bergmann, Kilian Batzner, Michael Fauser, David Sattlegger, and Carsten Steger. 2022. Beyond dents and scratches: Logical constraints in unsupervised anomaly detection and localization. International Journal of Computer Vision 130, 4 (2022), 947–969

2022

[3] [3]

Paul Bergmann, Michael Fauser, David Sattlegger, and Carsten Steger. 2019. MVTec AD–A comprehensive real-world dataset for unsupervised anomaly de- tection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 9592–9600

2019

[4] [4]

Paul Bergmann, Michael Fauser, David Sattlegger, and Carsten Steger. 2020. Uninformed students: Student-teacher anomaly detection with discriminative latent embeddings. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 4183–4192

2020

[5] [5]

Paul Bergmann, Xin Jin, David Sattlegger, and Carsten Steger. 2021. The mvtec 3d-ad dataset for unsupervised 3d anomaly detection and localization. arXiv preprint arXiv:2112.09045 (2021)

arXiv 2021

[6] [6]

Luca Bonfiglioli, Marco Toschi, Davide Silvestri, Nicola Fioraio, and Daniele De Gregorio. 2022. The eyecandies dataset for unsupervised multimodal anomaly detection and localization. In Proceedings of the Asian Conference on Computer Vision. 3586–3602

work page 2022

[7] [7]

Yunkang Cao, Xiaohao Xu, and Weiming Shen. 2024. Complementary pseudo multimodal feature for point cloud anomaly detection. Pattern Recognition 156 (2024), 110761

work page 2024

[8] [8]

Qiyu Chen, Huiyuan Luo, Chengkan Lv, and Zhengtao Zhang. 2024. A unified anomaly synthesis strategy with gradient ascent for industrial anomaly detection and localization. arXiv preprint arXiv:2407.09359 (2024)

arXiv 2024