Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

MetaUAS: Universal Anomaly Segmentation with One-Prompt Meta-Learning

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 A model trained only on synthetic before/after image pairs can segment never-seen anomalies from one normal image prompt.

desk verdict A genuinely useful pure-vision one-prompt anomaly segmentation model trained on synthetic change pairs, but the 'any unseen anomaly' claim outruns the per-class and Goods results. read the letter →

arxiv 2505.09265 v1 pith:INHXR6WA submitted 2025-05-14 cs.CV cs.AI

Bin-Bin Gao This is my paper · ORCID
classification cs.CVcs.AI
keywords anomalysegmentationchangemeta-learningone-promptzero-shotfeaturealignmentsyntheticdataindustrialdefectdetection
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

This paper tries to establish that anomaly segmentation—finding the pixels where a manufactured object deviates from normal—can be solved by a pure vision model that has never seen an anomaly or an anomaly dataset. It unifies anomaly segmentation into change segmentation: treat the one normal image as the 'before' picture and the query image as the 'after' picture, so every defect is just a changed region. The training data is synthetic, made by erasing objects from everyday images, pasting objects into them, and filling random patches with foreign pixels; each pair comes with an exact change mask. The authors claim that a model trained this way, using a one-prompt meta-learning setup with a soft feature alignment module, transfers to real industrial defects in new object categories, outperforming previous zero-shot, few-shot, and even full-shot methods while using only one normal image prompt and no language. If true, universal anomaly segmentation would no longer depend on vision-language models or on collecting target anomalies.

What carries the argument

The machinery is a three-part comparison: a frozen multi-scale encoder extracts features from prompt and query; a soft feature alignment module replaces each query feature vector with a softmax-weighted combination of prompt feature vectors across spatial locations, so mild geometric misalignment does not break the comparison; and a U-Net decoder fuses the aligned prompt features with query features to produce a pixel-level change heatmap. The training signal is the binary cross-entropy between the predicted heatmap and the known synthetic change mask. This is the piece that bridges paired-image change perception and single-image semantic segmentation.

What would settle it

Run the same training recipe on synthetic pairs that include only object-level inpainting and pasting, with no local Perlin fills; then measure per-defect-type pixel recall on MVTec classes with thin scratches and micro-cracks, such as screw and transistor. If the model segments coarse changes well but misses nearly all fine-grained defects while a model trained on the target anomalies succeeds, the synthetic-change distribution is insufficient and the universal claim fails.

Watch

Extended reading notes

Core claim

The central claim, stated as a rethinking of the task, is that anomaly segmentation is a special case of change segmentation: anomalies are exactly 'appearance', 'disappearance', or 'exchange' changes between a normal prompt image and a query image of the same class. On this basis the paper builds MetaUAS, a frozen-encoder network whose feature alignment module compares prompt and query features at multiple scales, using soft (softmax-weighted) alignment of prompt features toward query positions, and whose decoder emits a pixel-level change heatmap. The model is trained only on synthetic image pairs synthesized from ordinary images and instance masks—objects made to disappear by inpainting, objects pasted in, and Perlin-noise local fills—so the supervision is a change mask, not an anomaly annotation. Evaluated directly on three industrial anomaly benchmarks without any target-domain training, the authors report that the one-prompt model matches or beats zero-shot vision-language methods and full-shot unified anomaly segmentation models, with a much smaller and faster network.

Load-bearing premise

The load-bearing premise is that synthetic changes—erasing whole objects, pasting objects in, and filling Perlin-noise-shaped regions with pixels from other images—are representative enough of real-world anomalies that a model trained to spot them will also spot scratches, dents, and micro-cracks in object classes it has never seen.

Editorial extensions

If this is right

  • A pure vision model can serve as a universal anomaly segmenter without vision-language guidance, using one normal image prompt and no target-domain training.
  • The same model can be trained without any anomalous images: only large-scale synthetic change pairs with accurate masks are needed.
  • At inference the model is training-free and fast, with roughly an order of magnitude fewer parameters and a hundred-fold speed-up relative to a language-guided baseline, while still beating it on pixel-level precision-recall.
  • Prompt quality is a controllable factor: using the best-matched normal image as the prompt substantially improves over a random normal prompt, especially when classes contain many subcategories, as on the 484-subcategory Goods benchmark.
  • Adding a CLIP-derived visual prior pushes segmentation accuracy higher still, but the gains are additive; the core result stands without it.

Reading between the lines

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

  • The paper does not test this, but if change segmentation is the true substrate, then any paired images with known change masks—surveillance frames, multi-date satellite shots, before/after repair photos—should be interchangeable training data for the same recipe.
  • A likely blind spot the paper does not report: per-defect-type breakdowns would probably show the weakest transfer on thin scratches and micro-cracks, because the synthetic distribution is dominated by coarse object-level and large-patch changes; that is a prediction, not a paper claim.
  • The large jump between random and best-matched prompts on the many-subcategory benchmark suggests prompt selection is doing real work; a testable extension is to make the model prompt-robust through multi-prompt alignment instead of cosine-similarity retrieval.
  • The one-prompt formulation could transfer to medical imaging: a normal scan as prompt and the patient scan as query, with the model flagging any changed tissue as an anomaly, since the mechanism is purely visual comparison.
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.

Referee Report

3 major / 5 minor

Summary. The manuscript proposes MetaUAS, a pure-vision framework for universal anomaly segmentation. Anomaly segmentation is reformulated as change segmentation between a normal prompt image and a query image. A model with a frozen EfficientNet-b4 encoder, a soft feature alignment module, and a U-Net decoder is trained on synthetic image pairs derived from MS-COCO: object-level disappearance/appearance via inpainting, exchange via random pasting, and local-region changes via Perlin-mask filling. At inference, one normal image prompt is supplied per class (randomly selected in the base protocol, best-matched in the starred variant). Experiments on MVTec, VisA, and Goods report image- and pixel-level metrics, ablations, and efficiency comparisons. The authors claim state-of-the-art zero-/few-shot and even full-shot performance with a single normal prompt.

Significance. The core idea is interesting and the random-prompt protocol is honest: on MVTec, MetaUAS reaches P-PR 59.3 versus 38.4 for WinCLIP+, and its efficiency advantage is clearly demonstrated. The training distribution is disjoint from the evaluation benchmarks, anomaly masks are not used during training, and the ablations are systematic; the promised code and models also strengthen reproducibility. However, the headline claim is broader than the evidence: per-class results show weak segmentation on fine-grained defects, and the 'outperforms full-shot' statement is metric-dependent. The paper would be a solid empirical contribution if the claims were aligned with the evidence.

major comments (3)
  1. [Table A1 and Section 3.3] Table A1 shows that random-prompt MetaUAS has P-PR of only 8.1 on screw, 23.6 on capsule, 37.2 on transistor, and 25.1 on grid, while classes dominated by large surface defects such as carpet, tile, and wood exceed 70. This pattern is exactly what one would expect from a model trained on coarse object-level and Perlin-blob changes, and it contradicts the abstract's claim that the model segments 'any novel or unseen visual anomalies.' The authors should provide per-defect-type evidence (for example, scratches, dents, micro-cracks) or restrict the universality claim to anomaly types covered by the synthetic distribution.
  2. [Table 1 and Section 4.2 Generalization] The statement that MetaUAS 'significantly outperforms previous zero-shot, few-shot, and even full-shot anomaly segmentation methods' is not supported across all metrics. On MVTec, UniAD achieves higher I-ROC (96.7 vs 90.7) and P-PRO (90.0 vs 82.6); on VisA, UniAD leads I-ROC (90.8 vs 81.2). MetaUAS leads mainly on P-PR and P-F1max, which are important but do not justify a global full-shot superiority claim. The text should be revised to metric-specific and benchmark-specific conclusions.
  3. [Table 1 (Goods) and Section 5 Limitation] On Goods, random-prompt MetaUAS is near chance (I-ROC 54.5, P-PR 8.6), and the strong result (I-ROC 90.1) is obtained only by MetaUAS*, which retrieves the best-matched prompt from the full normal training pool. The Limitations section acknowledges prompt sensitivity but does not state that the headline universal result on Goods depends on access to the full normal set, rather than the single-prompt protocol. This distinction is load-bearing for the contribution claimed in the abstract and Section 3.2.
minor comments (5)
  1. [Section 3.2] In the Overview paragraph, 'contacted' should be 'concatenated'.
  2. [Section 3.4] The subsection heading contains a typo: 'Traning' should be 'Training'.
  3. [Table 1] The dataset name is misspelled as 'Gooods' in the table header; it should be 'Goods'.
  4. [Section 1] The sentence referring to the 'metal model' appears to be a typo for 'mental model'.
  5. [Section 4.2 Efficiency] The claim of '10× fewer parameters and 100× speed improvement' for MetaUAS* relative to WinCLIP+ is imprecise when compared with Table 2: at 256x256, MetaUAS* uses 22.1M versus 339.5M parameters (about 15x fewer) and 3.1 ms versus 201.3 ms (about 65x faster).

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the synthetic training pairs and industrial benchmarks are disjoint, and the one-prompt change-segmentation reformulation is a conceptual reduction rather than an equation-level tautology.

full rationale

MetaUAS trains a change-segmentation model on synthetic MS-COCO image pairs with object-level inpainting/pasting and Perlin-mask local fills (Section 3.3), then evaluates on MVTec, VisA, and Goods without using target anomaly masks or fitting any parameter to target data. The central reduction 'anomaly segmentation = change segmentation between a normal prompt and a query' (Section 3.1) is a definitional reformulation: change segmentation is defined independently, and its training supervision comes from synthetic masks, not from anomaly labels. No equation is reused as a prediction, and no fitted constant is relabeled as a result. The only self-citation [60] supporting the appearance/disappearance/exchange trichotomy is not load-bearing, because the trichotomy is stated informally and the same synthesis procedure follows external works (CYWS [50] and DRAEM [72]). The MetaUAS* variant retrieves a normal prompt from the target-domain normal pool, which is a protocol choice acknowledged in the Limitation section and is not circular reasoning; per-class weaknesses on screw, capsule, and grid are generalization and correctness concerns outside the circularity definition. Consequently no specific circular step can be quoted, and the score reflects only a negligible self-citation trace.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the synthetic-data distribution, the frozen backbone, and the alignment module. None of these are derived from first principles; the paper's evidence is entirely empirical benchmark performance, and the key domain assumptions remain untested on failure modes.

free parameters (5)
  • object-vs-local change mixing probability = 0.5
    During training, object-level and local-region changes are applied with probability 0.5 (Appendix A); Table 3c shows both help, but the exact ratio is hand-chosen.
  • number of synthetic training pairs = 60000
    Synthesized from 60,000 MS-COCO images with 0.95:0.05 train-validation split (Appendix A); Table 3d shows performance grows with data, so the scale is a chosen knob.
  • input resolution = 256x256
    Main results use 256x256 (Table 1); Table 2 shows 512x512 gives similar or slightly lower scores, so resolution is a hand-picked trade-off.
  • backbone = EfficientNet-b4 frozen
    Table 3b compares E-b4, E-b6, EViT-b3, MobileNetV2 and finds E-b4 frozen best; this is an empirical choice, not derived.
  • optimizer hyperparameters = lr 1e-4, wd 5e-4, 30 epochs
    Appendix A reports these values; no sensitivity analysis is provided.
assumptions (4)
  • domain assumption Anomaly segmentation reduces to change segmentation when the prompt is a normal image of the same class as the query.
    Section 3.1 equates anomalies with appearance, disappearance, and exchange changes; this equivalence is the foundation of the training scheme and is motivated by predictive coding but not proven.
  • domain assumption Synthetic changes derived from MS-COCO object masks are diverse enough to induce a universal change detector that transfers to industrial defects.
    Section 3.3 assumes object-level and local-region synthesis spans the space of real anomalies; only aggregate benchmark scores are offered as support.
  • domain assumption Features from an ImageNet-pretrained encoder generalize to industrial images without degradation.
    Section 3.2 freezes EfficientNet-b4; if ImageNet features cannot separate normal from defective industrial regions, the comparison-based approach degrades.
  • domain assumption Soft alignment via cross-similarity softmax can compensate for geometric variation between prompt and query.
    Equations 5 and 6 implement the alignment; Table 3a shows it helps empirically, but classes with large pose changes such as screw still have low P-PR, so the assumption is imperfect.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MetaUAS: Universal Anomaly Segmentation with One-Prompt Meta-Learning." pith.science (2026). https://pith.science/paper/INHXR6WA

@misc{pith2026250509265,
  author       = {Pith},
  title        = {Pith review of: MetaUAS: Universal Anomaly Segmentation with One-Prompt Meta-Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/INHXR6WA}},
  note         = {Machine review of arXiv:2505.09265}
}
read the original abstract

Zero- and few-shot visual anomaly segmentation relies on powerful vision-language models that detect unseen anomalies using manually designed textual prompts. However, visual representations are inherently independent of language. In this paper, we explore the potential of a pure visual foundation model as an alternative to widely used vision-language models for universal visual anomaly segmentation. We present a novel paradigm that unifies anomaly segmentation into change segmentation. This paradigm enables us to leverage large-scale synthetic image pairs, featuring object-level and local region changes, derived from existing image datasets, which are independent of target anomaly datasets. We propose a one-prompt Meta-learning framework for Universal Anomaly Segmentation (MetaUAS) that is trained on this synthetic dataset and then generalizes well to segment any novel or unseen visual anomalies in the real world. To handle geometrical variations between prompt and query images, we propose a soft feature alignment module that bridges paired-image change perception and single-image semantic segmentation. This is the first work to achieve universal anomaly segmentation using a pure vision model without relying on special anomaly detection datasets and pre-trained visual-language models. Our method effectively and efficiently segments any anomalies with only one normal image prompt and enjoys training-free without guidance from language. Our MetaUAS significantly outperforms previous zero-shot, few-shot, and even full-shot anomaly segmentation methods. The code and pre-trained models are available at https://github.com/gaobb/MetaUAS.

Figures

Figures reproduced from arXiv: 2505.09265 by the authors.

Figure 1
Figure 1. The proposed MetaUAS consists of an encoder, a feature alignment module (FAM), and a decoder. It is trained on a synthesized dataset in a one-prompt meta-learning manner for change segmentation tasks. Once trained, it can segment any anomalies providing only one normal image prompt. 3 Method 3.1 Rethinking Anomaly Segmentation Unlike traditional image segmentation, since anomaly appearance has various ways, it is ha… view at source ↗
Figure 2
Figure 2. Selected synthesizing image pairs and their change masks. (a) and (b) simulate “appearance” and “disappearance” synthesizing with mask inpainting [56], and [50], (c) simulate “exchange” synthesizing with random pasting, and (d) simulate local region changes synthesizing with DRAEM [72]. Object-Level Change. In the famous MS￾COCO [35], instances are annotated with poly￾gons, and thus their foreground masks are avail￾… view at source ↗
Figure 3
Figure 3. Qualitative comparisons with state-of-the-art methods on MVTec, VisA and Goods. In both two sub-figures (left and right), (b) and (g) represent query images and their anomaly masks, while (a) represent the corresponding normal image prompts. The predicted anomaly maps are shown using different methods, including (c) WinCLIP+ [26], (d) AnomalyCLIP [76], (e) UniAD [70] and (f) our MetaUAS. Best viewed in color and zoo… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Anomaly segmentation for query images with different normal image prompts including 5 random prompts and the optimal prompt (denoting as prompt⋆). The anomaly segmentation maps are generated with MetaUAS, MetaUAS⋆ and MetaUAS⋆+. 4.3 Ablation Study We perform component-…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Closed form perturbative relativistic modifications to wave-packet dynamics in the quantum harmonic oscillator

    quant-ph 2026-03 unverdicted novelty 5.0 of 10

    Closed-form O(1/c²) relativistic corrections to QHO wave-packet widths, variances, and uncertainty products leave minimum-uncertainty saturation intact and become percent-level for 1–10 keV electron confinement.

Reference graph

Works this paper leans on

78 extracted references · 70 canonical work pages · cited by 1 Pith paper

  1. [1]

    Street-view change detection with deconvolutional networks.Autonomous Robots, 42, 2018

    Pablo F Alcantarilla, Simon Stent, German Ros, Roberto Arroyo, and Riccardo Gherardi. Street-view change detection with deconvolutional networks.Autonomous Robots, 42, 2018

    work page 2018

  2. [2]

    Generalized denoising auto- encoders as generative models

    Yoshua Bengio, Li Yao, Guillaume Alain, and Pascal Vincent. Generalized denoising auto- encoders as generative models. InNeurIPS, 2013

    work page 2013

  3. [3]

    Beyond dents and scratches: Logical constraints in unsupervised anomaly detection and localization

    Paul Bergmann, Kilian Batzner, Michael Fauser, David Sattlegger, and Carsten Steger. Beyond dents and scratches: Logical constraints in unsupervised anomaly detection and localization. IJCV, 130(4), 2022

    work page 2022

  4. [4]

    MVTec-AD: A compre- hensive real-world dataset for unsupervised anomaly detection

    Paul Bergmann, Michael Fauser, David Sattlegger, and Carsten Steger. MVTec-AD: A compre- hensive real-world dataset for unsupervised anomaly detection. InCVPR, 2019

    work page 2019

  5. [5]

    Uninformed Students: Student-teacher anomaly detection with discriminative latent embeddings

    Paul Bergmann, Michael Fauser, David Sattlegger, and Carsten Steger. Uninformed Students: Student-teacher anomaly detection with discriminative latent embeddings. InCVPR, 2020

    work page 2020

  6. [6]

    Efficientvit: Lightweight multi-scale attention for high-resolution dense prediction

    Han Cai, Junyan Li, Muyan Hu, Chuang Gan, and Song Han. Efficientvit: Lightweight multi-scale attention for high-resolution dense prediction. InICCV, 2023

    work page 2023

  7. [7]

    A spatial-temporal attention-based method and a new dataset for remote sensing image change detection.Remote Sensing, 12(10), 2020

    Hao Chen and Zhenwei Shi. A spatial-temporal attention-based method and a new dataset for remote sensing image change detection.Remote Sensing, 12(10), 2020

    work page 2020

  8. [8]

    A closer look at the training strategy for modern meta-learning

    Jiaxin Chen, Xiao-Ming Wu, Yanke Li, Qimai Li, Li-Ming Zhan, and Fu-lai Chung. A closer look at the training strategy for modern meta-learning. InNeurIPS, 2020

    work page 2020

Show all 78 references
  1. [9]

    Deep one-class classification via interpolated gaussian descriptor

    Yuanhong Chen, Yu Tian, Guansong Pang, and Gustavo Carneiro. Deep one-class classification via interpolated gaussian descriptor. InAAAI, 2022

    2022

  2. [10]

    Self-supervised normalizing flows for image anomaly detection and localization

    Li-Ling Chiu and Shang-Hong Lai. Self-supervised normalizing flows for image anomaly detection and localization. InICCV, 2023

    2023

  3. [11]

    The relationship between precision-recall and roc curves

    Jesse Davis and Mark Goadrich. The relationship between precision-recall and roc curves. In ICML, 2006

    2006

  4. [12]

    PaDiM: a patch distribution modeling framework for anomaly detection and localization

    Thomas Defard, Aleksandr Setkov, Angelique Loesch, and Romaric Audigier. PaDiM: a patch distribution modeling framework for anomaly detection and localization. InICPR, 2021

    2021

  5. [13]

    Anomaly detection via reverse distillation from one-class embed- ding

    Hanqiu Deng and Xingyu Li. Anomaly detection via reverse distillation from one-class embed- ding. InCVPR, 2022

    2022

  6. [14]

    Catching both gray and black swans: Open-set supervised anomaly detection

    Choubo Ding, Guansong Pang, and Chunhua Shen. Catching both gray and black swans: Open-set supervised anomaly detection. InCVPR, 2022

    2022

  7. [15]

    FastRecon: Few-shot industrial anomaly detection via fast feature reconstruction

    Zheng Fang, Xiaoyang Wang, Haocheng Li, Jiejie Liu, Qiugui Hu, and Jimin Xiao. FastRecon: Few-shot industrial anomaly detection via fast feature reconstruction. InICCV, 2023

    2023

  8. [16]

    Distributed hierarchical processing in the primate cerebral cortex.Cerebral cortex, 1(1), 1991

    Daniel J Felleman and David C Van Essen. Distributed hierarchical processing in the primate cerebral cortex.Cerebral cortex, 1(1), 1991

    1991

  9. [17]

    Learning to detect multi-class anomalies with just one normal image prompt

    Bin-Bin Gao. Learning to detect multi-class anomalies with just one normal image prompt. In ECCV, 2024

    2024

  10. [18]

    Anomaly detection in video via self-supervised and multi-task learning

    Mariana-Iuliana Georgescu, Antonio Barbalau, Radu Tudor Ionescu, Fahad Shahbaz Khan, Marius Popescu, and Mubarak Shah. Anomaly detection in video via self-supervised and multi-task learning. InCVPR, 2021

    2021

  11. [19]

    Simple copy-paste is a strong data augmentation method for instance segmentation

    Golnaz Ghiasi, Yin Cui, Aravind Srinivas, Rui Qian, Tsung-Yi Lin, Ekin D Cubuk, Quoc V Le, and Barret Zoph. Simple copy-paste is a strong data augmentation method for instance segmentation. InCVPR, 2021

    2021

  12. [20]

    Memorizing normality to detect anomaly: Memory- augmented deep autoencoder for unsupervised anomaly detection

    Dong Gong, Lingqiao Liu, Vuong Le, Budhaditya Saha, Moussa Reda Mansour, Svetha Venkatesh, and Anton van den Hengel. Memorizing normality to detect anomaly: Memory- augmented deep autoencoder for unsupervised anomaly detection. InICCV, 2019. 11

    2019

  13. [21]

    Anoma- lyGPT: Detecting industrial anomalies using large vision-language models

    Zhaopeng Gu, Bingke Zhu, Guibo Zhu, Yingying Chen, Ming Tang, and Jinqiao Wang. Anoma- lyGPT: Detecting industrial anomalies using large vision-language models. InAAAI, 2024

    2024

  14. [22]

    Few-shot anomaly-driven generation for anomaly classification and segmentation

    Guan Gui, Bin-Bin Gao, Jun Liu, Chengjie Wang, and Yunsheng Wu. Few-shot anomaly-driven generation for anomaly classification and segmentation. InECCV, 2024

    2024

  15. [23]

    Divide- and-Assemble: Learning block-wise memory for unsupervised anomaly detection

    Jinlei Hou, Yingying Zhang, Qiaoyong Zhong, Di Xie, Shiliang Pu, and Hong Zhou. Divide- and-Assemble: Learning block-wise memory for unsupervised anomaly detection. InICCV, 2021

    2021

  16. [24]

    Registration based few-shot anomaly detection

    Chaoqin Huang, Haoyan Guan, Aofan Jiang, Ya Zhang, Michael Spratling, and Yan-Feng Wang. Registration based few-shot anomaly detection. InECCV, 2022

    2022

  17. [25]

    Decoupled weight decay regularization

    Loshchilov Ilya and Hutter Frank. Decoupled weight decay regularization. InICLR, 2019

    2019

  18. [26]

    WinCLIP: Zero-/few-shot anomaly classification and segmentation

    Jongheon Jeong, Yang Zou, Taewan Kim, Dongqing Zhang, Avinash Ravichandran, and Onkar Dabeer. WinCLIP: Zero-/few-shot anomaly classification and segmentation. InCVPR, 2023

    2023

  19. [27]

    SoftPatch: Unsupervised anomaly detection with noisy data

    Xi Jiang, Jianlin Liu, Jinbao Wang, Qiang Nie, Kai WU, Yong Liu, Chengjie Wang, and Feng Zheng. SoftPatch: Unsupervised anomaly detection with noisy data. InNeurIPS, 2022

    2022

  20. [28]

    Feasibility of anomaly score detected with deep learning in irradiated breast cancer patients with reconstruction.npj Digit

    Dong-Yun Kim, Soo Jin Lee, Eun-Kyu Kim, Eunyoung Kang, Chan Yeong Heo, Jae Hoon Jeong, Yujin Myung, In Ah Kim, and Bum-Sup Jang. Feasibility of anomaly score detected with deep learning in irradiated breast cancer patients with reconstruction.npj Digit. Med., 5(1), 2022

    2022

  21. [29]

    PyramidFlow: High-resolution defect contrastive localization using pyramid normalizing flow

    Jiarui Lei, Xiaobo Hu, Yue Wang, and Dong Liu. PyramidFlow: High-resolution defect contrastive localization using pyramid normalizing flow. InCVPR, 2023

    2023

  22. [30]

    Zero-shot anomaly detection via batch normalization

    Aodong Li, Chen Qiu, Marius Kloft, Padhraic Smyth, Maja Rudolph, and Stephan Mandt. Zero-shot anomaly detection via batch normalization. InNeurIPS, 2023

    2023

  23. [31]

    CutPaste: Self-supervised learning for anomaly detection and localization

    Chun-Liang Li, Kihyuk Sohn, Jinsung Yoon, and Tomas Pfister. CutPaste: Self-supervised learning for anomaly detection and localization. InCVPR, 2021

    2021

  24. [32]

    PromptAD: Learning prompts with only normal samples for few-shot anomaly detection

    Xiaofan Li, Zhizhong Zhang, Xin Tan, Chengwei Chen, Yanyun Qu, Yuan Xie, and Lizhuang Ma. PromptAD: Learning prompts with only normal samples for few-shot anomaly detection. InCVPR, 2024

    2024

  25. [33]

    MuSc: Zero-shot industrial anomaly classification and segmentation with mutual scoring of the unlabeled images

    Xurui Li, Ziming Huang, Feng Xue, and Yu Zhou. MuSc: Zero-shot industrial anomaly classification and segmentation with mutual scoring of the unlabeled images. InICLR, 2024

    2024

  26. [34]

    Feature pyramid networks for object detection

    Tsung-Yi Lin, Piotr Dollár, Ross Girshick, Kaiming He, Bharath Hariharan, and Serge Belongie. Feature pyramid networks for object detection. InCVPR, 2017

    2017

  27. [35]

    Microsoft COCO: Common objects in context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C Lawrence Zitnick. Microsoft COCO: Common objects in context. InECCV, 2014

    2014

  28. [36]

    Diversity-measurable anomaly detection

    Wenrui Liu, Hong Chang, Bingpeng Ma, Shiguang Shan, and Xilin Chen. Diversity-measurable anomaly detection. InCVPR, 2023

    2023

  29. [37]

    SimpleNet: A simple network for image anomaly detection and localization

    Zhikang Liu, Yiming Zhou, Yuansheng Xu, and Zilei Wang. SimpleNet: A simple network for image anomaly detection and localization. InCVPR, 2023

    2023

  30. [38]

    Inter-realization channels: Unsupervised anomaly detection beyond one-class classification

    Declan McIntosh and Alexandra Branzan Albu. Inter-realization channels: Unsupervised anomaly detection beyond one-class classification. InICCV, 2023

    2023

  31. [39]

    VT-ADL: A vision transformer network for image anomaly detection and localization

    Pankaj Mishra, Riccardo Verk, Daniele Fornasier, Claudio Piciarelli, and Gian Luca Foresti. VT-ADL: A vision transformer network for image anomaly detection and localization. InISIE, 2021

    2021

  32. [40]

    Learning memory-guided normality for anomaly detection

    Hyunjong Park, Jongyoun Noh, and Bumsub Ham. Learning memory-guided normality for anomaly detection. InCVPR, 2020. 12

    2020

  33. [41]

    OCGAN: One-class novelty detection using GANs with constrained latent representations

    Pramuditha Perera, Ramesh Nallapati, and Bing Xiang. OCGAN: One-class novelty detection using GANs with constrained latent representations. InCVPR, 2019

    2019

  34. [42]

    Learning transferable visual models from natural language supervision

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervision. InICML, 2021

    2021

  35. [43]

    Predictive coding in the visual cortex: a functional interpretation of some extra-classical receptive-field effects.Nature Neuroscience, 2(1), 1999

    Rajesh PN Rao and Dana H Ballard. Predictive coding in the visual cortex: a functional interpretation of some extra-classical receptive-field effects.Nature Neuroscience, 2(1), 1999

    1999

  36. [44]

    Modeling the distribution of normal data in pretrained deep features for anomaly detection

    Oliver Rippel, Patrick Mertens, and Dorit Merhof. Modeling the distribution of normal data in pretrained deep features for anomaly detection. InICPR, 2021

    2021

  37. [45]

    Self-supervised predictive convolutional attentive block for anomaly detection

    Nicolae-C˘at˘alin Ristea, Neelu Madan, Radu Tudor Ionescu, Kamal Nasrollahi, Fahad Shahbaz Khan, Thomas B Moeslund, and Mubarak Shah. Self-supervised predictive convolutional attentive block for anomaly detection. InCVPR, 2022

    2022

  38. [46]

    U-Net: Convolutional networks for biomedical image segmentation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-Net: Convolutional networks for biomedical image segmentation. InMICCAI, 2015

    2015

  39. [47]

    Towards total recall in industrial anomaly detection

    Karsten Roth, Latha Pemula, Joaquin Zepeda, Bernhard Schölkopf, Thomas Brox, and Peter Gehler. Towards total recall in industrial anomaly detection. InCVPR, 2022

    2022

  40. [48]

    Same same but differnet: Semi- supervised defect detection with normalizing flows

    Marco Rudolph, Bastian Wandt, and Bodo Rosenhahn. Same same but differnet: Semi- supervised defect detection with normalizing flows. InWACV, 2021

    2021

  41. [49]

    Fully convolutional cross-scale-flows for image-based defect detection

    Marco Rudolph, Tom Wehrbein, Bodo Rosenhahn, and Bastian Wandt. Fully convolutional cross-scale-flows for image-based defect detection. InWACV, 2022

    2022

  42. [50]

    The change you want to see

    Ragav Sachdeva and Andrew Zisserman. The change you want to see. InWACV, 2023

    2023

  43. [51]

    Learning to detect everything in an open world

    Kuniaki Saito, Ping Hu, Trevor Darrell, and Kate Saenko. Learning to detect everything in an open world. InECCV, 2022

    2022

  44. [52]

    Multiresolution knowledge distillation for anomaly detection

    Mohammadreza Salehi, Niousha Sadjadi, Soroosh Baselizadeh, Mohammad H Rohban, and Hamid R Rabiee. Multiresolution knowledge distillation for anomaly detection. InCVPR, 2021

    2021

  45. [53]

    MobileNetV2: Inverted residuals and linear bottlenecks

    Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, and Liang-Chieh Chen. MobileNetV2: Inverted residuals and linear bottlenecks. InCVPR, 2018

    2018

  46. [54]

    A hierarchical transformation-discriminating generative model for few shot anomaly detection

    Shelly Sheynin, Sagie Benaim, and Lior Wolf. A hierarchical transformation-discriminating generative model for few shot anomaly detection. InICCV, 2021

    2021

  47. [55]

    Real-world anomaly detection in surveillance videos

    Waqas Sultani, Chen Chen, and Mubarak Shah. Real-world anomaly detection in surveillance videos. InCVPR, 2018

    2018

  48. [56]

    Resolution-robust large mask inpainting with fourier convolutions

    Roman Suvorov, Elizaveta Logacheva, Anton Mashikhin, Anastasia Remizova, Arsenii Ashukha, Aleksei Silvestrov, Naejin Kong, Harshith Goka, Kiwoong Park, and Victor Lempitsky. Resolution-robust large mask inpainting with fourier convolutions. InWACV, 2022

    2022

  49. [57]

    EfficientNet: Rethinking model scaling for convolutional neural networks

    Mingxing Tan and Quoc Le. EfficientNet: Rethinking model scaling for convolutional neural networks. InICML, 2019

    2019

  50. [58]

    Inferotemporal cortex and object vision.Annual review of neuroscience, 19(1), 1996

    Keiji Tanaka. Inferotemporal cortex and object vision.Annual review of neuroscience, 19(1), 1996

    1996

  51. [59]

    Extracting and composing robust features with denoising autoencoders

    Pascal Vincent, Hugo Larochelle, Yoshua Bengio, and Pierre-Antoine Manzagol. Extracting and composing robust features with denoising autoencoders. InICML, 2008

    2008

  52. [60]

    How to reduce change detection to semantic segmentation.PR, 2023

    Guo-Hua Wang, Bin-Bin Gao, and Chengjie Wang. How to reduce change detection to semantic segmentation.PR, 2023

    2023

  53. [61]

    Student-teacher feature pyramid matching for anomaly detection.BMVC, 2021

    Guodong Wang, Shumin Han, Errui Ding, and Di Huang. Student-teacher feature pyramid matching for anomaly detection.BMVC, 2021. 13

    2021

  54. [62]

    Glancing at the patch: Anomaly localization with global and local feature comparison

    Shenzhi Wang, Liwei Wu, Lei Cui, and Yujun Shen. Glancing at the patch: Anomaly localization with global and local feature comparison. InCVPR, 2021

    2021

  55. [63]

    DatasetDM: Synthesizing data with perception annotations using diffusion models

    Weijia Wu, Yuzhong Zhao, Hao Chen, Yuchao Gu, Rui Zhao, Yefei He, Hong Zhou, Mike Zheng Shou, and Chunhua Shen. DatasetDM: Synthesizing data with perception annotations using diffusion models. InNeurIPS, 2023

    2023

  56. [64]

    DiffuMask: Synthesizing images with pixel-level annotations for semantic segmentation using diffusion models

    Weijia Wu, Yuzhong Zhao, Mike Zheng Shou, Hong Zhou, and Chunhua Shen. DiffuMask: Synthesizing images with pixel-level annotations for semantic segmentation using diffusion models. InICCV, 2023

    2023

  57. [65]

    In-painting radiography images for unsupervised anomaly detection

    Tiange Xiang, Yongyi Lu, Alan L Yuille, Chaoyi Zhang, Weidong Cai, and Zongwei Zhou. In-painting radiography images for unsupervised anomaly detection. InCVPR, 2023

    2023

  58. [66]

    Pushing the limits of fewshot anomaly detection in industry vision: Graphcore

    Guoyang Xie, Jingbao Wang, Jiaqi Liu, Feng Zheng, and Yaochu Jin. Pushing the limits of fewshot anomaly detection in industry vision: Graphcore. InICLR, 2023

    2023

  59. [67]

    Learning semantic context from normal samples for unsupervised anomaly detection

    Xudong Yan, Huaidong Zhang, Xuemiao Xu, Xiaowei Hu, and Pheng-Ann Heng. Learning semantic context from normal samples for unsupervised anomaly detection. InAAAI, 2021

    2021

  60. [68]

    Focus the Discrepancy: Intra-and inter-correlation learning for image anomaly detection

    Xincheng Yao, Ruoqi Li, Zefeng Qian, Yan Luo, and Chongyang Zhang. Focus the Discrepancy: Intra-and inter-correlation learning for image anomaly detection. InICCV, 2023

    2023

  61. [69]

    Explicit boundary guided semi-push-pull contrastive learning for supervised anomaly detection

    Xincheng Yao, Ruoqi Li, Jing Zhang, Jun Sun, and Chongyang Zhang. Explicit boundary guided semi-push-pull contrastive learning for supervised anomaly detection. InCVPR, 2023

    2023

  62. [70]

    A unified model for multi-class anomaly detection

    Zhiyuan You, Lei Cui, Yujun Shen, Kai Yang, Xin Lu, Yu Zheng, and Xinyi Le. A unified model for multi-class anomaly detection. InNeurIPS, 2022

    2022

  63. [71]

    Old is Gold: Redefining the adversarially learned one-class classifier training paradigm

    Muhammad Zaigham Zaheer, Jin-ha Lee, Marcella Astrid, and Seung-Ik Lee. Old is Gold: Redefining the adversarially learned one-class classifier training paradigm. InCVPR, 2020

    2020

  64. [72]

    DRAEM: A discriminatively trained reconstruction embedding for surface anomaly detection

    Vitjan Zavrtanik, Matej Kristan, and Danijel Sko ˇcaj. DRAEM: A discriminatively trained reconstruction embedding for surface anomaly detection. InICCV, 2021

    2021

  65. [73]

    Reconstruction by inpainting for visual anomaly detection.PR, 112, 2021

    Vitjan Zavrtanik, Matej Kristan, and Danijel Skoˇcaj. Reconstruction by inpainting for visual anomaly detection.PR, 112, 2021

    2021

  66. [74]

    PKU-GoodsAD: A supermarket goods dataset for unsupervised anomaly detection and segmentation.RA-L, 9(3), 2024

    Jian Zhang, Runwei Ding, Miaoju Ban, and Linhui Dai. PKU-GoodsAD: A supermarket goods dataset for unsupervised anomaly detection and segmentation.RA-L, 9(3), 2024

    2024

  67. [75]

    OmniAL: A unified cnn framework for unsupervised anomaly localization

    Ying Zhao. OmniAL: A unified cnn framework for unsupervised anomaly localization. In CVPR, 2023

    2023

  68. [76]

    AnomalyCLIP: Object- agnostic prompt learning for zero-shot anomaly detection

    Qihang Zhou, Guansong Pang, Yu Tian, Shibo He, and Jiming Chen. AnomalyCLIP: Object- agnostic prompt learning for zero-shot anomaly detection. InICLR, 2024

    2024

  69. [77]

    Toward generalist anomaly detection via in-context residual learning with few-shot sample prompts

    Jiawen Zhu and Guansong Pang. Toward generalist anomaly detection via in-context residual learning with few-shot sample prompts. InCVPR, 2024

    2024

  70. [78]

    A photo of a normal [cls]

    Yang Zou, Jongheon Jeong, Latha Pemula, Dongqing Zhang, and Onkar Dabeer. Spot-the- difference self-supervised pre-training for anomaly detection and segmentation. InECCV, 2022. 14 (a) (b) (c) (d) (e) (f) (g) (a) (b) (c) (d) (e) (f) (g) Figure A1:Qualitative comparisons with s...

    2022

Pith tools

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