Pith. sign in

REVIEW 5 major objections 5 minor 89 references

Understanding and Overcoming Cross-modal Fusion Bias in Multimodal Anomaly Detection From A Fisher Information Perspective

T0 review · 5 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read This paper claims that cross-modal fusion bias in RGB-depth anomaly detection is a measurable imbalance in each encoder's Fisher information trace, and that rebalancing it early with a plug-in module (UCFB) improves existing MAD methods.

desk verdict A useful plug-in for MAD with a shaky Fisher-information story; the empirical gains are modest but real, the causal claims need reining in. read the letter →

arxiv 2608.00986 v1 pith:EQJ6VD5L submitted 2026-08-02 cs.CV

classification cs.CV
keywords multimodalanomalydetectioncross-modalfusionbiasFisherinformationmatrixmodalitysuppressionRGB-depthunimodaladaptiveadjustmentcanonicalsimilarityanalysisfew-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

This paper sets out to show that cross-modal fusion bias—one sensor's encoder out-learning and silencing the other—is a real, diagnosable bottleneck in multimodal anomaly detection, not a footnote. Using the trace of the Fisher Information Matrix as a per-modality measure of information acquisition, the authors find that RGB and depth encoders acquire information at comparable rates when trained alone but diverge sharply when trained jointly, with the imbalance locked in during the earliest epochs. Five blur-based ablation cases show that disturbing this early phase permanently lowers fusion performance, while later disturbances barely matter. The proposed UCFB module detects the dominant modality and critical window, slows the dominant encoder with a Fisher-guided regularizer, and strengthens the suppressed encoder through canonical similarity analysis. Plugged into four existing MAD baselines, it yields consistent I-AUROC gains on MVTec 3D-AD and Eyecandies across single-class, multi-class, and few-shot settings while keeping memory and speed comparable.

What carries the argument

The central object is the per-modality Fisher information trace, defined as $Tr(F^\kappa) = \mathbb{E}_{x_\kappa \sim \chi_\kappa}\left[\|\nabla_{w_\kappa} \mathcal{L}(w_\kappa, x_\kappa)\|^2\right]$, which the paper treats as each modality's information-acquisition level. It is used three times: to compute the dominance gap $\Delta_\kappa$ that names the dominant modality, to detect the critical learning epoch via the relative drop of the trace between consecutive epochs against a threshold $\Upsilon$, and to set the regularization strength $\delta = \exp(\mu \tanh(\Delta_\kappa))$ for a squared-parameter-change penalty $\Gamma^t_{\kappa,b}$ that slows the dominant encoder. The second mecha

What would settle it

Construct a variant of MVTec 3D-AD where the depth channel is replaced by random noise while RGB carries all anomaly signal. If the Fisher trace gap between RGB and depth still appears in the early epochs and equalizing it with UCFB still improves I-AUROC, then the gap is not specifically about task-relevant suppression; if the gain disappears, the suppression story is confirmed.

Watch

Extended reading notes

Core claim

The paper's central claim is that cross-modal fusion bias in RGB-depth anomaly detection is visible in the trace of each modality's Fisher Information Matrix (the expected squared norm of that modality's encoder gradients). Under joint training, one trace rises much faster than the other in the first epochs; once this dominance is established, the suppressed modality never recovers, and blur experiments indicate that early interference is irreversible. From this, the paper derives a design principle: interventions must occur in the critical early learning epoch, slowing the dominant modality while boosting the suppressed one. UCFB implements this with unimodal adaptive adjustment—a Fisher-gu

Load-bearing premise

The load-bearing premise is that the Fisher information trace faithfully measures task-relevant information acquisition, so a large RGB-depth trace gap during joint training reflects one modality genuinely suppressing the other (and that blurring an input simulates that suppression); if the gap is a generic training artifact, the diagnosis and the UCFB design do not follow.

Editorial extensions

If this is right

  • Existing MAD methods can be upgraded by inserting UCFB without redesigning their fusion modules, since it operates on per-modality gradients and features.
  • The first training epochs are decisive for fusion quality; damage done there is not repaired by simply training longer, so future methods should schedule any balancing intervention early.
  • The Fisher trace gap can serve as a live, cheap diagnostic for detecting when one modality is being silenced, because it is computable from gradients already available during training.
  • Re-balancing information acquisition—slowing the dominant modality and strengthening the suppressed one—is a viable alternative to inventing new fusion architectures for anomaly detection.
  • The gains hold even with very limited data (5/10/50-shot), so the module does not require large training sets to be useful.

Reading between the lines

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

  • Editorial extension: the same Fisher trace diagnostic could be applied to other multimodal tasks such as RGB-thermal tracking or video-audio recognition, where early modality suppression is also suspected; the paper tests only RGB-depth anomaly detection.
  • Editorial extension: the paper's critical-epoch finding implies that simple curriculum choices—like warming up the weaker modality first or ordering training data—might reproduce part of UCFB's gain; this is a cheap experiment the paper does not run.
  • Editorial extension: because the reported gains are averages over categories, an informative follow-up would be to check whether UCFB's benefit concentrates in categories with intrinsically weak depth signal, which would sharpen the causal story.
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

5 major / 5 minor

Summary. The paper studies cross-modal fusion bias in RGB-depth anomaly detection. It first measures modality-specific information acquisition by the squared gradient norm (trace of the Fisher Information Matrix) and claims that multimodal training induces one modality suppressing the other. It then presents ablative experiments using blurred inputs to argue that early-stage interference is irreversibly harmful, and identifies a 'critical learning epoch.' Based on this analysis, the authors propose UCFB, a plug-and-play module combining unimodal adaptive adjustment (UAA), which slows the dominant modality via Fisher-information-guided regularization, and canonical similarity analysis (CSA), which improves cross-modal interaction. The method is integrated into four MAD baselines and evaluated on MVTec 3D-AD and Eyecandies in single-class, multi-class, and few-shot settings, reporting small average improvements over the baselines.

Significance. If the central analysis were correct, this would be one of the first works to explicitly diagnose and mitigate cross-modal fusion bias in multimodal anomaly detection, and the plug-and-play nature of UCFB would make the contribution practically attractive. The empirical comparisons do show gains on two standard benchmarks, and the paper includes useful ablations and visualizations. However, the conceptual contribution rests on two unvalidated premises: that the FIM-trace gap is a faithful measure of task-relevant information suppression, and that input blurring replicates the effect of that suppression. Because these premises motivate the UAA design, the 'understanding' contribution is not established. The empirical gains are also mostly small and reported without statistical significance, making the strength of the performance claim uncertain.

major comments (5)
  1. [Section 2.1, Eq. (1), Fig. 1] The quantity Tr(F_kappa) defined in Eq. (1) is the expected squared norm of the gradient of the total loss with respect to the modality encoder's parameters. Since the total loss is computed after fusion, this gradient contains contributions backpropagated through the fusion module and depends on the other modality's branch. Thus a small trace does not necessarily mean the modality is suppressed; it may indicate that the features are already predictable from the other modality or that the encoder has a different effective parameterization. Observation 1 compares unimodal training (separate heads) with per-modality traces in a multimodal model, so the observed gap can reflect architectural change rather than inter-modal competition. The UAA design and the paper's 'understanding' claim rest on interpreting this gap as suppression. Please validate the measure against an independent modality
  2. [Section 2.2, Table 1, Cases 1–5] The ablative cases replace one modality with a blurred input. Blurring removes information at the input level; it does not reproduce the gradient-level out-competition that UAA later modulates. These experiments show that early input corruption is harmful and irreversible, but they do not establish that the FIM-trace imbalance observed in Fig. 1 is the causal mechanism of suboptimal fusion. The design of UAA follows directly from this causal premise, so a direct manipulation of the hypothesized mechanism is needed (e.g., scaling one modality's gradient norm in early training, or ablating the UAA module and measuring both the FIM-trace gap and performance jointly). As presented, the connection between the blur interventions and the method is asserted rather than validated.
  3. [Tables 2–5] All quantitative comparisons report a single run per method/setting. Several improvements are small (0.1–0.3% I-AUROC in Tables 4–5). Without multiple seeds, confidence intervals, or significance tests, the claim of 'consistent improvements' is not supported. Please provide mean ± std over at least 3 seeds (or paired statistical tests) for the main comparisons, including the few-shot results.
  4. [Section 4.3, Table 6] UCFB has two hyperparameters, Υ (Eq. 4) and μ (Eq. 12), and Table 6 shows strong sensitivity on EasyNet (I-AUROC ranging from 91.7 to 93.8). The main experiments (Tables 2–5) do not state which values were used, nor whether the same values were applied across all baselines, datasets, and settings. If Υ and μ were tuned on the benchmark itself, the reported gains may partly reflect hyperparameter selection rather than the method. Report the chosen values and evaluate sensitivity on a validation split.
  5. [Section 3.3, Eq. (6) and Lemma 1] The approximation Γ^t_{κ,b} ≈ (δξ²/2) Σ_i ||g^t_{ψκ,i}||² relies on Lemma 1, which assumes gradients are random vectors drawn uniformly from a high-dimensional sphere. Real mini-batch gradients are correlated through the data and the optimization trajectory, and they are not uniformly distributed on the sphere. The paper's Remark acknowledges this is idealized, but an empirical check of the cross-term magnitude is not provided. If the cross terms are not negligible, Γ^t_{κ,b} does not approximate Tr(F^t_{κ,b}) as claimed. Please add empirical evidence or an alternative justification.
minor comments (5)
  1. [Section 2.1, Eq. (1)] The notation is inconsistent: the equation defines g_φκ but the trace uses g_ψκ. Please unify the symbol for the modality encoder parameters.
  2. [Section 3.3, Lemma 2] The proof of Lemma 2 is said to be in the Supplementary Materials. Since this is a convergence claim used to justify the method, either include the proof or state the exact assumptions and a proof sketch in the main text.
  3. [Table 1] The table lists 'Baseline+UCFB (Ours)' for CFM and EasyNet, but the corresponding numbers (96.1 and 93.8) appear in later tables. It would be clearer to refer to the main experimental section and avoid duplicating results without context.
  4. [Figure 4] The captions say 'FIM trace' while the y-axis labels say 'Information acquisition amount.' Please align the terminology and state the units (e.g., squared gradient norm) for reproducibility.
  5. [Overall] The paper claims 'theoretical results' in the contributions, but the only theoretical content is Lemma 1 (an idealized orthogonality claim) and Lemma 2 (deferred). If the authors intend a theoretical contribution, they should clarify which precise statements are proven and under what assumptions.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the performance claims rest on external benchmarks, and the self-citations and hyperparameter choices are not load-bearing.

full rationale

The paper's central performance claim—that UCFB improves MAD baselines—is evaluated against external benchmarks (MVTec 3D-AD and Eyecandies) and published baseline methods, so it does not reduce to a fitted constant or to the paper's own definitions. The Fisher-information trace in Eq. (1) is a defined monitoring quantity, and UAA directly regularizes it, but the claimed MAD improvement is measured independently by I-AUROC/P-AUROC/AUPRO. The hyperparameters Υ and μ are tuned on the benchmark in Table 6, but this is standard model selection rather than a 'prediction' that equals its fit. The CSA module is attributed to an external source [29] as well as to the authors' prior work [46], so the self-citation is not the sole support. References [45] and [47] are also self-citations but are used as baseline/existing work, not as an unverified uniqueness or foundational theorem. The blur ablations and FIM-trace analysis support the motivation but are inferential; that is an evidence-strength concern, not a circularity. No equation in the paper is shown to be equivalent to its own output by construction, and no fitted input is renamed as a prediction.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The method rests on two hand-tuned hyperparameters (Upsilon, mu) and two strong modeling assumptions: that squared gradient norms measure information acquisition, and that blurring simulates suppression. The orthogonality lemma is a standard heuristic but is applied to gradients in a way that is not strictly justified. No new physical entities are introduced.

free parameters (2)
  • threshold Upsilon (Eq. 4) = 0.01 or 0.04 in ablations; not stated for main results
    Determines whether the model is in a critical period; tuned per-dataset (Table 6), no separate validation set described.
  • regularization strength mu (Eq. 12) = 0.9 best in Table 6; values 0.1, 0.5, 0.9 explored
    Controls delta = exp(mu * tanh(delta_kappa)); no principled choice; tuned on the same benchmark.
assumptions (4)
  • domain assumption Gradients from different mini-batches are approximately orthogonal in high-dimensional spaces (Lemma 1, Eq. 7-11), so cross terms in the FIM-trace approximation vanish.
    Used to derive Gamma in Eq. 6; real gradients are correlated, so the approximation is heuristic; the paper admits it is an idealized assumption in the Remark after Lemma 1.
  • domain assumption The squared gradient norm (trace of FIM) is a faithful measure of modality-specific information acquisition.
    Equates information acquired with gradient magnitude; deeper models may have higher gradient norms without contributing more task-relevant signal.
  • ad hoc to paper Blurring an input modality replicates the effect of cross-modal suppression in the ablative cases of Section 2.2.
    Blurring destroys information rather than simulating a gradient-competition process; the equivalence to modality suppression is not established.
  • domain assumption Lemma 2 convergence-rate preservation holds under sufficiently small delta and xi; proof deferred to supplementary.
    The precise conditions are not in the main text, so the convergence guarantee is unverified in the paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Understanding and Overcoming Cross-modal Fusion Bias in Multimodal Anomaly Detection From A Fisher Information Perspective." pith.science (2026). https://pith.science/paper/EQJ6VD5L

@misc{pith2026260800986,
  author       = {Pith},
  title        = {Pith review of: Understanding and Overcoming Cross-modal Fusion Bias in Multimodal Anomaly Detection From A Fisher Information Perspective},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EQJ6VD5L}},
  note         = {Machine review of arXiv:2608.00986}
}
read the original abstract

Current advancements in Multimodal Anomaly Detection (MAD) are largely driven by enhancing multimodal fusion, particularly through the integration of RGB and Depth data for richer anomaly representation. However, less attention was devoted to analyzing the role of cross-modal fusion bias, a well-known challenge in multimodal learning, in MAD. This gap motivates a key question: can we overcome this bias to break the performance bottleneck of current work? In this paper, we first analyze the impact of cross-modal fusion bias in MAD via the Fisher Information Matrix. Then, grounded in these findings, we propose UCFB, a simple yet effective plug-and-play framework designed to mitigate cross-modal fusion bias in MAD. It achieves this by jointly employing Fisher-information-guided dynamic calibration to adjust modality-specific regularization weights and canonical similarity analysis to improve inter-modal interactions. Extensive experiments on the MVTec 3D-AD and Eyecandies datasets demonstrate that UCFB achieves consistent improvements in single-class, multi-class, and few-shot settings.

Figures

Figures reproduced from arXiv: 2608.00986 by the authors.

Figure 1
Figure 1. The gap in the trace of FIM between the RGB and depth modalities for three baseline models under unimodal [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. (Top) Visualization of different ablative cases; [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Overview of UCFB. During training, UCFB first identifies the dominant modality (Step 1), checks whether it is in the [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: (Left) The value of FIM trace for the RGB and depth [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

89 extracted references · 73 canonical work pages

  1. [1]

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

  2. [2]

    Paul Bergmann, Xin Jin, David Sattlegger, and Carsten Steger. 2022. The MVTec 3D-AD Dataset for Unsupervised 3D Anomaly Detection and Localization. In Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications. 202–213

  3. [3]

    Paul Bergmann and David Sattlegger. 2023. Anomaly detection in 3d point clouds using deep geometric descriptors. InProceedings of the W ACV. 2613–2623

  4. [4]

    Luca Bonfiglioli, Marco Toschi, Davide Silvestri, Nicola Fioraio, and Daniele De Gregorio. 2022. The eyecandies dataset for unsupervised multimodal anomaly detection and localization. InProceedings of the ACCV. 3586–3602

  5. [5]

    Yunkang Cao, Jiangning Zhang, Luca Frittoli, Yuqi Cheng, Weiming Shen, and Giacomo Boracchi. 2024. Adaclip: Adapting clip with hybrid learnable prompts for zero-shot anomaly detection. InEuropean Conference on Computer Vision. 55–72

  6. [6]

    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. InProceedings of the 31st ACM International Conference on Multimedia. 7038–7046

  7. [7]

    Jiayi Cheng, Can Gao, Jie Zhou, Jiajun Wen, Tao Dai, and Jinbao Wang. 2025. MC3D-AD: A Unified Geometry-aware Reconstruction Model for Multi-category 3D Anomaly Detection. InProceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence. 837–845

  8. [8]

    Yuqi Cheng, Yunkang Cao, Haiming Yao, Wei Luo, Cheng Jiang, Hui Zhang, and Weiming Shen. 2026. A comprehensive survey for real-world industrial surface defect detection: Challenges, approaches, and prospects.Journal of Manufacturing Systems84 (2026), 152–172

Show all 89 references
  1. [9]

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

  2. [10]

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

  3. [11]

    Zhewei Dai, Shilei Zeng, Haotian Liu, Xurui Li, Feng Xue, and Yu Zhou. 2025. Seas: Few-shot industrial anomaly image generation with separation and sharing fine-tuning. InProceedings of the IEEE/CVF International Conference on Computer Vision. 23135–23144

  4. [12]

    Thomas Defard, Aleksandr Setkov, Angelique Loesch, and Romaric Audigier

  5. [13]

    Hanqiu Deng and Xingyu Li. 2022. Anomaly detection via reverse distillation from one-class embedding. InProceedings of the IEEE/CVF conference on CVPR. 9737–9746

  6. [14]

    Ronald Aylmer Fisher. 1925. Theory of statistical estimation. InMathematical proceedings of the Cambridge philosophical society, Vol. 22. 700–725

  7. [15]

    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. InProceedings of the IEEE/CVF International Co...

  8. [16]

    Zhihao Gu, Jiangning Zhang, Liang Liu, Xu Chen, Jinlong Peng, Zhenye Gan, Guannan Jiang, Annan Shu, Yabiao Wang, and Lizhuang Ma. 2024. Rethinking reverse distillation for multi-modal anomaly detection. InProceedings of the AAAI Conference on Artificial Intelligence, Vol. 38. ...

  9. [17]

    Haoyang He, Jiangning Zhang, Hongxu Chen, Xuhai Chen, Zhishan Li, Xu Chen, Yabiao Wang, Chengjie Wang, and Lei Xie. 2024. A diffusion-based framework for multi-class anomaly detection. InProceedings of the AAAI conference on artificial intelligence, Vol. 38. 8472–8480

  10. [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 CVPR. 2968–2977

  11. [19]

    Teng Hu, Jiangning Zhang, Ran Yi, Yuzhen Du, Xu Chen, Liang Liu, Yabiao Wang, and Chengjie Wang. 2024. Anomalydiffusion: Few-shot anomaly image generation with diffusion model. InProceedings of the AAAI conference on artificial intelligence, Vol. 38. 8526–8534

  12. [20]

    Chaoqin Huang, Aofan Jiang, Jinghao Feng, Ya Zhang, Xinchao Wang, and Yanfeng Wang. 2024. Adapting visual-language models for generalizable anomaly detection in medical images. InProceedings of the IEEE/CVF Conference on CVPR. 11375–11385

  13. [21]

    Chengxiang Huang, Yake Wei, Zequn Yang, and Di Hu. 2025. Adaptive unimodal regulation for balanced multimodal information acquisition. InProceedings of the Computer Vision and Pattern Recognition Conference. 25854–25863

  14. [22]

    Xi Jiang, Jian Li, Hanqiu Deng, Yong Liu, Bin-Bin Gao, Yifeng Zhou, Jialin Li, Chengjie Wang, and Feng Zheng. 2025. MMAD: A Comprehensive Benchmark for Multimodal Large Language Models in Industrial Anomaly Detection. InThe Thirteenth International Conference on Learning Repre...

  15. [23]

    Xi Jiang, Jianlin Liu, Jinbao Wang, Qiang Nie, Kai Wu, Yong Liu, Chengjie Wang, and Feng Zheng. 2022. Softpatch: Unsupervised anomaly detection with noisy data.Advances in Neural Information Processing Systems35 (2022), 15433–15445

  16. [24]

    Soopil Kim, Sion An, Philip Chikontwe, Myeongkyun Kang, Ehsan Adeli, Kil- ian M Pohl, and Sang Hyun Park. 2024. Few shot part segmentation reveals compositional logic for industrial anomaly detection. InProceedings of the AAAI conference on artificial intelligence, Vol. 38. 8591–8599

  17. [25]

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

  18. [26]

    Hanxi Li, Jianfei Hu, Bo Li, Hao Chen, Yongbin Zheng, and Chunhua Shen. 2024. Target before shooting: Accurate anomaly detection and localization under one millisecond via cascade patch retrieval.IEEE Transactions on Image Processing (2024)

  19. [27]

    Hao Li, Yuhao Wang, Wenning Hao, Pingping Zhang, Dong Wang, and Huchuan Lu. 2026. RAGTrack: Language-aware RGBT Tracking with Retrieval-Augmented Generation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 28179–28189

  20. [28]

    Hao Li, Yuhao Wang, Xiantao Hu, Wenning Hao, Pingping Zhang, Dong Wang, and Huchuan Lu. 2026. Cadtrack: Learning contextual aggregation with de- formable alignment for robust rgbt tracking. InProceedings of the AAAI Conference on Artificial Intelligence, Vol. 40. 6109–6117

  21. [29]

    Po-han Li, Sandeep P Chinchali, et al. 2025. CSA: Data-efficient Mapping of Uni- modal Features to Multimodal Features. InThe Thirteenth International Conference on Learning Representations

  22. [30]

    Wenqiao Li, Xiaohao Xu, Yao Gu, Bozhong Zheng, Shenghua Gao, and Yingna Wu. 2024. Towards scalable 3d anomaly detection and localization: A benchmark via 3d anomaly synthesis and a self-supervised learning network. InProceedings of the IEEE/CVF conference on CVPR. 22207–22216

  23. [31]

    Wenqiao Li, Bozhong Zheng, Xiaohao Xu, Jinye Gan, Fading Lu, Xiang Li, Na Ni, Zheng Tian, Xiaonan Huang, Shenghua Gao, et al. 2025. Multi-sensor object anomaly detection: Unifying appearance, geometry, and internal properties. In Proceedings of the CVPR. 9984–9993

  24. [32]

    Xiaofan Li, Xin Tan, Zhuo Chen, Zhizhong Zhang, Ruixin Zhang, Rizen Guo, Guanna Jiang, Yulong Chen, Yanyun Qu, Lizhuang Ma, et al. 2025. One-for-more: Continual diffusion model for anomaly detection. InProceedings of the CVPR. 4766–4775

  25. [33]

    Yuqi Li, Junhao Dong, Jiao Liu, Piotr Koniusz, Hansheng Zeng, Chuanguang Yang, Junming Liu, Yingli Tian, Tingwen Huang, and Hao Wu. 2026. Evolving Multimodal Models for Physical Dynamics: A Multi-objective Neuroevolution Approach.IEEE Transactions on Evolutionary Computation(2026)

  26. [34]

    Yiting Li, Fayao Liu, Jingyi Liao, Sichao Tian, Chuan-Sheng Foo, and Xulei Yang

  27. [35]

    Yuqi Li, Chuanguang Yang, Hansheng Zeng, Zeyu Dong, Zhulin An, Yongjun Xu, Yingli Tian, and Hao Wu. 2025. Frequency-aligned knowledge distillation for lightweight spatiotemporal forecasting. InProceedings of the IEEE/CVF Inter- national Conference on Computer Vision. 7262–7272

  28. [36]

    Yu Xin Li, Felix Dangel, Derek Tam, and Colin Raffel. 2025. Fishers for Free? Approximating the Fisher Information Matrix by Recycling the Squared Gradient Accumulator. InForty-second International Conference on Machine Learning

  29. [37]

    Zixu Li, Yupeng Hu, Zhiwei Chen, Qinlei Huang, Guozhi Qiu, Zhiheng Fu, and Meng Liu. 2026. Retrack: Evidence-driven dual-stream directional anchor calibra- tion network for composed video retrieval. InProceedings of the AAAI Conference on Artificial Intelligence, Vol. 40. 23373–23381

  30. [38]

    Zixu Li, Yupeng Hu, Zhiwei Chen, Haokun Wen, Xuemeng Song, and Liqiang Nie. 2026. COMBINER: Composed Image Retrieval Guided by Attribute-based Neighbor Relations.IEEE Transactions on Image Processing(2026)

  31. [39]

    Zixu Li, Yupeng Hu, Zhiwei Chen, Mingyu Zhang, Zhiheng Fu, and Liqiang Nie

  32. [40]

    Yuxuan Lin, Hanjing Yan, Xuan Tong, Yang Chang, Huanzhen Wang, Ziheng Zhou, Shuyong Gao, Yan Wang, and Wenqiang Zhang. 2025. Commonality in Few: Few-Shot Multimodal Anomaly Detection via Hypergraph-Enhanced Memory. arXiv preprint arXiv:2511.05966(2025)

  33. [41]

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

  34. [42]

    Yu Liu, Yanbing Liu, Fangfang Yuan, Cong Cao, Youbang Sun, Kun Peng, WeiZhuo Chen, Jianjun Li, and Zhiyuan Ma. 2026. OPERA: A Reinforcement Learning– Enhanced Orchestrated Planner-Executor Architecture for Reasoning-Oriented Multi-Hop Retrieval. InProceedings of the AAAI Confe...

  35. [43]

    Zhipu Liu and Lei Zhang. 2025. Multi-model Synergy Perception for Open-world Person Re-identification.IEEE Transactions on Circuits and Systems for Video Technology(2025)

  36. [44]

    Zhikang Liu, Yiming Zhou, Yuansheng Xu, and Zilei Wang. 2023. Simplenet: A simple network for image anomaly detection and localization. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition. 20402–20411

  37. [45]

    Kaifang Long, Lianbo Ma, Jiaqi Liu, Liming Liu, and Guoyang Xie. 2026. Towards an Incremental Unified Multimodal Anomaly Detection: Augmenting Multimodal Denoising From an Information Bottleneck Perspective. InProceedings of the IEEE/CVF Conference on Computer Vision and Patte...

  38. [46]

    Kaifang Long, Guoyang Xie, Lianbo Ma, Qing Li, Min Huang, Jianhui Lv, and Zhichao Lu. 2025. Enhancing Multimodal Learning via Hierarchical Fusion Architecture Search With Inconsistency Mitigation.IEEE Transactions on Image Processing(2025)

  39. [47]

    InEuropean Conference on Computer Vision

    Learning diffusion models for multi-view anomaly detection. InEuropean Conference on Computer Vision. Springer, 328–345. MM ’26, November 10–14, 2026, Rio de Janeiro, Brazil. Kaifang Long et al

  40. [48]

    Kaifang Long, Han Zhao, Zengzhen Shao, Yang Cao, Yanfang Geng, Yintai Sun, Weizhi Xu, and Hui Yu. 2023. Deep neural network with embedding fusion for Chinese named entity recognition.ACM Transactions on Asian and Low-Resource Language Information Processing22, 3 (2023), 1–16

  41. [49]

    Wenxin Ma, Xu Zhang, Qingsong Yao, Fenghe Tang, Chenxu Wu, Yingtai Li, Rui Yan, Zihang Jiang, and S Kevin Zhou. 2025. Aa-clip: Enhancing zero-shot anomaly detection via anomaly-aware clip. InProceedings of the CVPR. 4744–4754

  42. [50]

    Kai Mao, Ping Wei, Yiyang Lian, Yangyang Wang, and Nanning Zheng. 2025. Beyond Single-Modal Boundary: Cross-Modal Anomaly Detection through Visual Prototype and Harmonization. InProceedings of the CVPR. 9964–9973

  43. [51]

    Declan McIntosh and Alexandra Branzan Albu. 2024. Unsupervised, online and on-the-fly anomaly detection for non-stationary image distributions. InEuropean Conference on Computer Vision. Springer, 428–445

  44. [52]

    Bingchen Miao, Wenqiao Zhang, Juncheng Li, Wangyu Wu, Siliang Tang, Zhaocheng Li, Haochen Shi, Jun Xiao, and Yueting Zhuang. 2025. Robust Modality-Incomplete Anomaly Detection: A Modality-Instructive Framework with Benchmark. InProceedings of the 33rd ACM International Confere...

  45. [53]

    Kaifang Long, Guoyang Xie, Lianbo Ma, Jiaqi Liu, and Zhichao Lu. 2025. Revisiting multimodal fusion for 3D anomaly detection from an architectural perspective. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 39. 12273–12281

  46. [54]

    Marco Rudolph, Tom Wehrbein, Bodo Rosenhahn, and Bastian Wandt. 2022. Fully convolutional cross-scale-flows for image-based defect detection. InProceedings of the IEEE/CVF winter conference on applications of computer vision. 1088–1097

  47. [55]

    Marco Rudolph, Tom Wehrbein, Bodo Rosenhahn, and Bastian Wandt. 2023. Asymmetric student-teacher networks for industrial anomaly detection. InPro- ceedings of the IEEE/CVF W ACV. 2592–2602

  48. [56]

    Ylli Sadikaj, Hongkuan Zhou, Lavdim Halilaj, Stefan Schmid, Steffen Staab, and Claudia Plant. 2025. Multiads: Defect-aware supervision for multi-type anomaly detection and segmentation in zero-shot learning. InProceedings of the IEEE/CVF International Conference on Computer Vi...

  49. [57]

    Hannah M Schlüter, Jeremy Tan, Benjamin Hou, and Bernhard Kainz. 2022. Nat- ural synthetic anomalies for self-supervised anomaly detection and localization. InEuropean Conference on Computer Vision. Springer, 474–489

  50. [58]

    Luc PJ Sträter, Mohammadreza Salehi, Efstratios Gavves, Cees GM Snoek, and Yuki M Asano. 2024. Generalad: Anomaly detection across domains by attending to distorted features. InEuropean Conference on Computer Vision. 448–465

  51. [59]

    Xiaokang Peng, Yake Wei, Andong Deng, Dong Wang, and Di Hu. 2022. Balanced multimodal learning via on-the-fly gradient modulation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition. 8238–8247

  52. [60]

    Chengyu Tao, Xuanming Cao, and Juan Du. 2025. G2SF: Geometry-Guided Score Fusion for Multimodal Industrial Anomaly Detection. InProceedings of the IEEE/CVF International Conference on Computer Vision. 20551–20560

  53. [61]

    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. InProceedings of the IEEE/CVF conference on CVPR. 24511–24520

  54. [62]

    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. InEuropean Conference on Computer Vision. 75–91

  55. [63]

    Xiaohan Wang, Zhangtao Cheng, Ting Zhong, Leiting Chen, and Fan Zhou

  56. [64]

    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

  57. [65]

    Jiaqi Tang, Hao Lu, Xiaogang Xu, Ruizheng Wu, Sixing Hu, Tong Zhang, Tsz Wa Cheng, Ming Ge, Ying-Cong Chen, and Fugee Tsung. 2024. An incremental unified framework for small defect inspection. InEuropean conference on computer vision. 307–324

  58. [66]

    Jiawei Xu, Qiangqiang Zhou, Zhouping Li, Yanjiao Shi, Yugen Yi, and Jiacong Yu

  59. [67]

    Jiawei Xu, Qiangqiang Zhou, Dandan Zhu, Yong Chen, Yugen Yi, and Xiaoqi Zhao. 2026. TP-Seg: Task-Prototype Framework for Unified Medical Lesion Segmentation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 5452–5462

  60. [68]

    Jianan Ye, Weiguang Zhao, Xi Yang, Guangliang Cheng, and Kaizhu Huang. 2025. Po3ad: Predicting point offsets toward better 3d point cloud anomaly detection. In Proceedings of the Computer Vision and Pattern Recognition Conference. 1353–1362

  61. [69]

    Zhiyuan You, Lei Cui, Yujun Shen, Kai Yang, Xin Lu, Yu Zheng, and Xinyi Le

  62. [70]

    Modality-Balanced Collaborative Distillation for Multi-Modal Domain Generalization. InAAAI. 26535–26543

  63. [71]

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

  64. [72]

    Yake Wei, Di Hu, Henghui Du, and Ji-Rong Wen. 2025. On-the-Fly Modulation for Balanced Multimodal Learning.IEEE Transactions on Pattern Analysis and Machine Intelligence47, 1 (2025), 469–485

  65. [73]

    Lei Zhang, Zhipu Liu, Wensheng Zhang, and David Zhang. 2023. Style uncertainty based self-paced meta learning for generalizable person re-identification.IEEE Transactions on Image Processing32 (2023), 2107–2119

  66. [74]

    HVPNet: A Bio-Inspired Network for General Salient and Camouflaged Object Detection.Neural Networks(2026), 109340

  67. [75]

    Ximiao Zhang, Min Xu, and Xiuzhuang Zhou. 2024. Realnet: A feature selection network with realistic synthetic anomaly for anomaly detection. InProceedings of the IEEE/CVF conference on CVPR. 16699–16708

  68. [76]

    Zhe Zhang, Mingxiu Cai, Hanxiao Wang, Gaochang Wu, Tianyou Chai, and Xia- tian Zhu. 2025. CostFilter-AD: Enhancing Anomaly Detection through Matching Cost Filtering. InForty-second International Conference on Machine Learning

  69. [77]

    Qicheng Zhao, Yu Li, Qi Sun, and Zheyu Yan. 2026. ResilPhase: Plug-and-Play Phase Mapping and Noise-Resilient Macro-Trajectory Extrapolation for Diffusion Acceleration.arXiv preprint arXiv:2606.26769(2026)

  70. [78]

    Qicheng Zhao, Qi Sun, and Zheyu Yan. 2026. Seeing the End at Step Zero: Accel- erating Diffusion MLLMs via MLP Sparsity-Aware Truncation.arXiv preprint arXiv:2607.14557(2026)

  71. [79]

    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

  72. [80]

    Qiangqiang Zhou, Jiawei Xu, Yong Chen, Dandan Zhu, Yugen Yi, and Xiaoqi Zhao. 2026. DifferSeg: Towards Diverse Multimodal Binary Segmentation via Differential Perception and Frequency Guidance.IEEE Transactions on Circuits and Systems for Video Technology(2026)

  73. [81]

    Vitjan Zavrtanik, Matej Kristan, and Danijel Skočaj. 2024. Cheating depth: Enhancing 3d surface anomaly detection via depth simulation. InProceedings of the IEEE/CVF W ACV. 2164–2172

  74. [83]

    Wei Zhang, Yihang Wu, Shengkai Yu, Songhua Li, Qiang Li, and Qi Wang. 2026. GPR-MVS: Global Propagation Regularization for Large Scale Multi-view Stereo. IEEE Transactions on Geoscience and Remote Sensing(2026)

  75. [88]

    Bozhong Zheng, Jinye Gan, Xiaohao Xu, Xintao Chen, Wenqiao Li, Xiaonan Huang, Na Ni, and Yingna Wu. 2025. Bridging 3D Anomaly Localization and Repair via High-Quality Continuous Geometric Representation. InProceedings of the IEEE/CVF International Conference on Computer Vision...

  76. [2021]

    InInternational conference on pattern recognition

    Padim: a patch distribution modeling framework for anomaly detection and localization. InInternational conference on pattern recognition. Springer, 475– 489

  77. [2022]

    A unified model for multi-class anomaly detection.Advances in Neural Information Processing Systems35 (2022), 4571–4584

  78. [2023]

    InProceed- ings of the 40th International Conference on Machine Learning

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

  79. [2024]

    InProceedings of the IEEE/CVF Conference on CVPR

    Multimodal industrial anomaly detection by crossmodal feature mapping. InProceedings of the IEEE/CVF Conference on CVPR. 17234–17243

  80. [2025]

    InProceedings of the IEEE/CVF International Conference on Computer Vision

    FIND: Few-Shot Anomaly Inspection with Normal-Only Multi-Modal Data. InProceedings of the IEEE/CVF International Conference on Computer Vision. 23290–23299

  81. [2026]

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

    Conesep: Cone-based robust noise-unlearning compositional network for composed image retrieval. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 16897–16909

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.