Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

Video Anomaly Detection with Motion and Appearance Guided Patch Diffusion Model

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The paper claims that a patch-based conditional diffusion model, guided by appearance and motion, reaches state-of-the-art frame-level AUC on four video anomaly detection benchmarks.

desk verdict A plausible patch-diffusion recipe for VAD with strong numbers but a few evaluation loose ends and an untested vulnerability to anomaly copying through the appearance skip connections. read the letter →

arxiv 2412.09026 v1 pith:JPSOIPOS submitted 2024-12-12 cs.CV

Hang Zhou , Jiale Cai , Yuteng Ye , Yonghui Feng , Chenxing Gao , Junqing Yu , Zikai Song , Wei Yang This is my paper · ORCID
classification cs.CV
keywords videoanomalydetectiondiffusionmodelframepredictionpatch-basedprocessingappearanceandmotionconditionsmemorybanktemporaldifferenceDDIM
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

The paper tries to establish that a diffusion model operating on image patches, rather than whole frames or latent features, can detect video anomalies by predicting the next frame from the previous six. It argues that anomalies are often small local deviations in both what things look like and how they move, so the denoiser should be guided by an appearance condition (the first observed frame) and a motion condition (temporal differences between consecutive frames). Training only on normal videos, the model learns to generate normal continuations; at test time, anomalous content is flagged by a high mean-squared error between the predicted frame and the actual one. If correct, the method would beat existing prediction- and reconstruction-based detectors on Ped2, Avenue, ShanghaiTech, and UBnormal, reaching frame-level AUCs of 98.6, 91.3, 79.2, and 63.4 respectively.

What carries the argument

The central object is the patch-based conditional denoising network MA-PDM, which treats future-frame prediction as a conditional generation problem. Its load-bearing parts are: (1) random patch cropping during training and sliding-window patch merging with an overlap normalization matrix $R$ at inference, so diffusion runs on $64\times64$ patches instead of full $256\times256$ frames; (2) the appearance condition $F_a=F_1$ and the motion condition $F_m=F[:,-1]-F[:,1:]$, a temporal difference that is concatenated with the noisy frame as input to the noise-estimation U-Net; (3) a patch memory bank $M\in\mathbb{R}^{P\times N\times D}$ addressed by cross-attention to store normal semantics per patch location, with the filtered semantics embedded through adaptive layer normalization; and (4) the DDIM reverse process that assembles the predicted patch noises into a full-image noise map. This machinery carries the argument by making the denoiser able to recover only normal local patterns, so abnormal objects produce a high reconstruction error.

What would settle it

Run the trained MA-PDM on a test clip where the anomaly is already visible in the first observed frame and in the temporal-difference condition, and check whether the MSE anomaly score in Eq. 16 stays high or collapses; if the denoiser copies anomalous content from the conditioning frames, the detection margin would vanish exactly on such clips.

Watch

Extended reading notes

Core claim

The paper claims that the right unit for video anomaly detection is a local patch in pixel space, not a whole-frame or latent feature, and that a conditional diffusion model trained to denoise patches can be made into a top-performing frame-prediction detector by feeding it two complementary conditions: appearance, taken from the first of six observed frames, and motion, taken as the temporal difference between those frames. A learnable patch memory bank stores normal semantic patterns per patch location, and the filtered semantics are injected into every residual block through adaptive layer normalization, while the appearance encoder's hierarchical features are fused into the U-Net decoder through residual conditioning. At inference, Gaussian noise is denoised with DDIM over sliding-window patches, overlapping predictions are merged by an overlap normalization matrix, and the anomaly score is the MSE between the reconstructed future frame (plus the reconstructed first frame) and the ground truth. On four benchmarks the paper reports frame-level AUCs of 98.6 on Ped2, 91.3 on Avenue, 79.2 on ShanghaiTech, and 63.4 on UBnormal, and states that the approach outperforms all methods assessed.

Load-bearing premise

The method assumes that a model trained only on normal videos, conditioned on raw test frames that may themselves contain the anomaly, will still fail to generate the anomalous content, so the reconstruction error stays high.

Editorial extensions

If this is right

  • Replacing whole-image diffusion with $64\times64$ patch processing improves anomaly detection across datasets, and the improvement is larger when both the appearance and motion conditions are added on top.
  • Temporal difference is both faster and more effective than optical flow as a motion condition, with the paper reporting higher AUC and roughly double the frames per second.
  • Storing normal patterns per patch location in a memory bank outperforms a global memory bank, which supports the claim that local normality is the right granularity for surveillance anomalies.
  • The full MA-PDM model improves over the previous diffusion-based detector FPDM on Avenue, ShanghaiTech, and UBnormal, and matches or exceeds a masked-autoencoder baseline without using synthetic anomalies.
  • The method produces a working detector on UBnormal, a fully synthetic virtual-scene dataset, which the paper treats as a one-class anomaly detection test.

Reading between the lines

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

  • A direct testable extension would be to use the same appearance and motion conditions with classifier-free guidance; the paper does not explore guidance strength, which could trade normal-frame fidelity against anomaly sensitivity.
  • Because the anomaly score is the pixel-level MSE to the true future frame, the method implicitly assumes that every anomaly changes pixel content; slow or subtle anomalies that leave most pixels unchanged could evade detection even if the memory bank represents them.
  • The sliding-window patch predictions could be reused as a coarse anomaly heatmap, but the paper does not evaluate pixel-level localization, so any localization benefit from patch processing remains an unmeasured byproduct.
  • The ablation treats random cropping as data augmentation; if its main effect is dataset enlargement, the relative gain over whole-image diffusion may shrink on datasets with more varied scenes or non-static cameras.
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 paper proposes MA-PDM, a patch-based diffusion model for video anomaly detection within a future-frame-prediction framework. An appearance condition is taken from the first frame of a six-frame clip, a motion condition is defined as the temporal difference between consecutive frames, and a DDIM-based noise estimator is conditioned on both. A location-indexed patch memory bank constrains semantic embeddings from the appearance encoder, while hierarchical appearance features are injected into the decoder via ControlNet-style skip connections. At inference, sliding-window patches are denoised, merged, and the anomaly score is the MSE between the predicted and true future frame plus an alpha-weighted MSE between reconstructed and true appearance. Experiments on Ped2, Avenue, ShanghaiTech, and UBnormal report frame-level AUCs of 98.6, 91.3, 79.2, and 63.4, respectively, with ablations over modules, patch sizes, memory variants, and motion representations.

Significance. If the reported results are reproducible, MA-PDM is a competitive contribution: it outperforms the diffusion-based FPDM on Avenue, ShanghaiTech, and UBnormal, and is competitive with AED-MAE without synthetic anomalies. The patch-level formulation is well motivated for small anomalous objects, and the ablations provide evidence that each proposed component contributes to performance. The paper does not, however, provide code, error bars, or a validation protocol for the per-dataset hyperparameters, so the small reported margins cannot currently be assessed for robustness. The central assumption that conditioning on raw test frames does not leak anomalous content into the generated frame is also untested; this is the main risk to the validity of the anomaly score.

major comments (3)
  1. [Sec. 3.2, Eq. (4); Sec. 3.4; Sec. 3.5, Eq. (16)] The conditions used at inference are raw test frames: the appearance condition Fa = F1 and the motion condition Fm = F[:-1] - F[1:] are taken from the observed part of the test clip, which may itself contain the anomaly. The motion condition is concatenated channel-wise with the noisy target, and the hierarchical appearance features hA_e are injected directly into the decoder without passing through the patch memory bank. If the denoiser can copy anomalous content from these conditions, the MSE in Eq. (16) collapses and the reported AUCs no longer reflect detection of the anomaly. The paper neither states nor tests the assumption that conditioning on anomalous test frames will not help generate the anomalous target. This is load-bearing for the central claim, and the authors should add an explicit test, for example by comparing detection on anomalies that first appear only in the target frame versus anomalies already visible in the conditioning frames, or by ablating the conditions on anomalous frames.
  2. [Sec. 4.1 and Table 1] The anomaly-score weight alpha is set per dataset to (0, 0.2, 0.3, 1) and the training epochs are set per dataset to (1000, 300, 30, 40), but no validation procedure is described. Since alpha=0 on Ped2 means the appearance reconstruction term is completely discarded, and since the reported improvements over prior methods are small on several datasets, the headline numbers are a function of parameters selected on the test distributions. The authors should report a validation protocol, a sensitivity analysis over alpha and epochs, and ideally multiple runs with standard deviations. Without this, the claim that the method 'outperforms all other methods assessed' is not yet supported.
  3. [Table 1 and Sec. 4.2] The reported margins are sometimes very small (e.g., 79.2 vs. 79.1 on ShanghaiTech and a tie at 91.3 on Avenue), yet no error bars, number of runs, or statistical significance tests are provided. Frame-level AUC differences of this size can easily be within run-to-run noise, especially without a fixed random seed or code release. The authors should provide standard deviations over multiple training runs and, if possible, per-dataset significance tests, before claiming state-of-the-art status.
minor comments (5)
  1. [Sec. 3.4, Eq. (8)] The patch-location example is numerically inconsistent: for coordinates (58,58), p = (58/64 - 1/2)*4 + (58/64 - 1/2) yields approximately 2.03, not 5. Please correct the formula or the example.
  2. [Sec. 3.5 and Sec. 4.1] The sliding-window stride is described as r=32 in Sec. 3.5 but as 'a stride of 64' in Sec. 4.1. Please clarify which stride is used for the reported results.
  3. [Table 3 and Sec. 4.3] The text says 'our technique achieves the highest results on four datasets with a patch size of 64,' but Table 3 reports results for only Ped2, Avenue, and ShanghaiTech; no UBnormal results are shown for the patch-size ablation. Also, the text states that patch size 32 is more efficient for Ped2, but the table shows a tie at 98.6 between sizes 32 and 64 under random cropping.
  4. [Sec. 4.1] The training details say 'The batch size in all three datasets is 16' although four datasets are used; please correct this.
  5. [Throughout] There are several typos and unfinished textual references that should be cleaned up, including 'parampters' in Sec. 3.4 and the repeated sentence 'Detail framework is shown in Supplementary Material' with no supplementary material available in the arXiv version.

Circularity Check

1 steps flagged · score 6.0 of 10

Reported superiority is partly an artifact of per-dataset test-set tuning: the anomaly score's alpha weights (and patch size/epochs) are selected on the same test folds whose AUCs are then reported, so the 'prediction' is not independent of the benchmark labels.

  1. fitted input called prediction [Section 4.1 (Training Details) and Eq. (16); Section 4.2, Table 1]
    "The values forα are (0, 0.2, 0.3, 1) across four datasets. ... Score = MSE(IN +1, ˆIN +1) +αMSE(I1, ˆI1) (16)"

    Eq. 16 defines the anomaly score as MSE(IN+1, IN+1_hat) + α·MSE(I1, I1_hat). Section 4.1 fixes α per dataset to (0, 0.2, 0.3, 1) with no validation split or selection procedure described. Table 1 then reports frame-level AUC computed from these same scores on the same test sets. The reported 'prediction' is therefore a function of a free parameter chosen against the test distributions; the claim that MA-PDM 'outperforms all other methods' is not an independent forecast but an evaluation of a test-tuned configuration. The same holds for per-dataset training epochs (1000/300/30/40) and patch size chosen from test AUC in Table 3.

full rationale

The core derivation (Eqs. 1–15) is self-contained and uses standard DDIM/DDPM machinery; the memory-augmented appearance encoder is a normal reconstruction module, and citations to prior work for ControlNet/DAE-style components are not load-bearing. I do not find circularity in the equations themselves. The significant circular step is in the evaluation loop: Eq. 16's anomaly score contains a per-dataset weight α, and no validation procedure is reported; the same test folds are used both to select α (and patch size/epochs) and to report the AUCs in Table 1. Thus the headline result is partly a statement about a test-tuned configuration, not an independent prediction. The additional concern that anomalous test frames used as conditions could be copied through skip connections is a robustness/correctness assumption rather than a derivation-circularity step, so it is not scored here as circularity.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

Most of the machinery is standard DDPM and DDIM. The domain assumptions carry the weight: anomalies are local enough for 64x64 patches, normal-only training makes denoising error a reliable anomaly score, and conditioning on raw test frames does not contaminate the prediction. The patch memory bank is an ad hoc architectural choice validated only by an ablation. No physically invented entities are introduced.

free parameters (7)
  • anomaly score weight alpha per dataset = 0 (Ped2), 0.2 (Avenue), 0.3 (Shanghai), 1 (UBnormal)
    Section 4.1: "The values for alpha are (0, 0.2, 0.3, 1) across four datasets." No validation procedure is described; the final AUC depends directly on Eq.16.
  • patch size p = 64 for all datasets
    Table 3 ablation selects patch size on test AUC; the paper notes size 32 is more efficient for Ped2 but the final setup states 64x64 for all datasets (Section 4.1).
  • reconstruction loss weight lambda1 = 0.1
    Section 3.4 defines L = Lp + lambda1 * Lr with lambda1 = 0.1; no sensitivity analysis is provided.
  • DDIM reverse steps = 5
    Section 4.1 training details; chosen for speed, with no ablation on the number of reverse steps.
  • input clip length N = 6 observed frames
    Section 3.2 sets N=6 to predict frame 7; no ablation on clip length is reported.
  • training epochs per dataset = 1000 (Ped2), 300 (Avenue), 30 (Shanghai), 40 (UBnormal)
    Section 4.1 lists per-dataset epoch counts without a stated validation criterion, so they appear selected by test performance.
  • patch memory capacity = 16x64x256
    Section 4.1 fixes the memory capacity across all datasets; no ablation on memory size is reported.
assumptions (6)
  • standard math DDPM and DDIM forward and reverse diffusion equations define a valid denoising process.
    Used in Eq.1-2 and Eq.15; standard background from Ho et al. and Nichol and Dhariwal.
  • domain assumption Patches of size 64x64 capture local anomalies and preserve enough global context for detection.
    Section 3.3 justifies patch decomposition by arguing surveillance anomalies are small; no experiment varies anomaly scale independently.
  • domain assumption Training only on normal videos makes the denoiser reconstruct or predict normal patches with lower error than abnormal ones.
    This is the core generation-based VAD premise and is not verified directly; the paper reports frame-level AUC but not per-patch error separation.
  • domain assumption At test time, conditioning on a raw first frame F1 and temporal difference Fm that may contain anomalies does not cause the model to generate anomalous content.
    The appearance and motion conditions in Eq.4 are taken from the test clip unchanged; the paper does not analyze the case where the condition itself is abnormal.
  • domain assumption Temporal difference (frame subtraction) is a sufficient motion cue for separating normal and abnormal dynamics.
    Section 3.2 chooses TD over optical flow; Table 5 gives empirical support on two datasets, but no formal justification is provided.
  • ad hoc to paper The location-indexed patch memory bank can store normal semantics per grid cell without contamination from anomalous patches.
    Cross-attention to memory in Eq.8 is introduced for this architecture; no independent evidence shows the 16-cell partition is appropriate for all scenes.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Video Anomaly Detection with Motion and Appearance Guided Patch Diffusion Model." pith.science (2026). https://pith.science/paper/JPSOIPOS

@misc{pith2026241209026,
  author       = {Pith},
  title        = {Pith review of: Video Anomaly Detection with Motion and Appearance Guided Patch Diffusion Model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JPSOIPOS}},
  note         = {Machine review of arXiv:2412.09026}
}
read the original abstract

A recent endeavor in one class of video anomaly detection is to leverage diffusion models and posit the task as a generation problem, where the diffusion model is trained to recover normal patterns exclusively, thus reporting abnormal patterns as outliers. Yet, existing attempts neglect the various formations of anomaly and predict normal samples at the feature level regardless that abnormal objects in surveillance videos are often relatively small. To address this, a novel patch-based diffusion model is proposed, specifically engineered to capture fine-grained local information. We further observe that anomalies in videos manifest themselves as deviations in both appearance and motion. Therefore, we argue that a comprehensive solution must consider both of these aspects simultaneously to achieve accurate frame prediction. To address this, we introduce innovative motion and appearance conditions that are seamlessly integrated into our patch diffusion model. These conditions are designed to guide the model in generating coherent and contextually appropriate predictions for both semantic content and motion relations. Experimental results in four challenging video anomaly detection datasets empirically substantiate the efficacy of our proposed approach, demonstrating that it consistently outperforms most existing methods in detecting abnormal behaviors.

Figures

Figures reproduced from arXiv: 2412.09026 by the authors.

Figure 1
Figure 1. We propose a patch-based diffusion model with a [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Our MA-PDM comprises three components: a patch cropping module for creating patch conditions and noise, an [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Four examples of anomaly detection comparison on Ped2 and ShanghaiTech datasets. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

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. CA-Diff: Collaborative Anatomy Diffusion for Brain Tissue Segmentation

    eess.IV 2025-06 conditional novelty 6.0 of 10

    CA-Diff jointly denoises a brain label and an atlas-derived distance field with one diffusion U-Net, plus a consistency loss and time-adapted attention, and reports state-of-the-art Dice on MALC, SchizBull, and Hammers.

Reference graph

Works this paper leans on

47 extracted references · 36 canonical work pages · cited by 1 Pith paper

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Bao, Q.; Liu, F.; Liu, Y.; Jiao, L.; Liu, X.; and Li, L. 2022. Hierarchical scene normality-binding modeling for anomaly detection in surveillance videos. In Proceedings of the 30th ACM international conference on multimedia, 6103--6112

    work page 2022

  4. [4]

    Cai, R.; Zhang, H.; Liu, W.; Gao, S.; and Hao, Z. 2021. Appearance-motion memory consistency network for video anomaly detection. In Proceedings of the AAAI conference on artificial intelligence, volume 35, 938--946

    work page 2021

  5. [5]

    Chang, Y.; Tu, Z.; Xie, W.; and Yuan, J. 2020. Clustering driven deep autoencoder for video anomaly detection. In European Conference on Computer Vision, 329--345. Springer

    work page 2020

  6. [6]

    Chen, C.; Xie, Y.; Lin, S.; Yao, A.; Jiang, G.; Zhang, W.; Qu, Y.; Qiao, R.; Ren, B.; and Ma, L. 2022. Comprehensive regularization in a bi-directional predictive network for video anomaly detection. In Proceedings of the AAAI conference on artificial intelligence, volume 36, 230--238

    work page 2022

  7. [7]

    Couairon, G.; Verbeek, J.; Schwenk, H.; and Cord, M. 2022. Diffedit: Diffusion-based semantic image editing with mask guidance. arXiv preprint arXiv:2210.11427

    arXiv 2022

  8. [8]

    Dhariwal, P.; and Nichol, A. 2021. Diffusion models beat gans on image synthesis. In Advances in neural information processing systems, volume 34, 8780--8794

    work page 2021

Show all 47 references
  1. [9]

    R.; Venkatesh, S.; and Hengel, A

    Gong, D.; Liu, L.; Le, V.; Saha, B.; Mansour, M. R.; Venkatesh, S.; and Hengel, A. v. d. 2019. Memorizing normality to detect anomaly: Memory-augmented deep autoencoder for unsupervised anomaly detection. In Proceedings of the IEEE international conference on computer vision, ...

    2019

  2. [10]

    Ho, J.; Jain, A.; and Abbeel, P. 2020. Denoising diffusion probabilistic models. In Advances in neural information processing systems, 6840--6851

    2020

  3. [11]

    G.; Choi, D

    Lee, S.; Kim, H. G.; Choi, D. H.; Kim, H.-I.; and Ro, Y. M. 2021. Video prediction recalling long-term motion context via memory alignment learning. In Proceedings of the IEEE conference on computer vision and pattern recognition, 3054--3063

    2021

  4. [12]

    Liu, W.; Luo, W.; Lian, D.; and Gao, S. 2018. Future frame prediction for anomaly detection--a new baseline. In Proceedings of the IEEE conference on computer vision and pattern recognition, 6536--6545

    2018

  5. [13]

    Liu, Z.; Nie, Y.; Long, C.; Zhang, Q.; and Li, G. 2021. A hybrid video anomaly detection framework via memory-augmented flow reconstruction and flow-guided frame prediction. In Proceedings of the IEEE international conference on computer vision, 13588--13597

    2021

  6. [14]

    Lu, C.; Shi, J.; and Jia, J. 2013. Abnormal Event Detection at 150 FPS in MATLAB. In Proceedings of the IEEE international conference on computer vision, 2720--2727

    2013

  7. [15]

    Lu, Z.; Wu, C.; Chen, X.; Wang, Y.; Qiao, Y.; and Liu, X. 2023. Hierarchical Diffusion Autoencoders and Disentangled Image Manipulation. arXiv preprint arXiv:2304.11829

    2023 arXiv

  8. [16]

    Lugmayr, A.; Danelljan, M.; Romero, A.; Yu, F.; Timofte, R.; and Van Gool, L. 2022. Repaint: Inpainting using denoising diffusion probabilistic models. In Proceedings of the IEEE conference on computer vision and pattern recognition, 11461--11471

    2022

  9. [17]

    Luo, W.; Liu, W.; and Gao, S. 2017. A revisit of sparse coding based anomaly detection in stacked rnn framework. In Proceedings of the IEEE international conference on computer vision, 341--349

    2017

  10. [18]

    Lv, H.; Chen, C.; Cui, Z.; Xu, C.; Li, Y.; and Yang, J. 2021. Learning normal dynamics in videos with meta prototype network. In Proceedings of the IEEE conference on computer vision and pattern recognition, 15425--15434

    2021

  11. [19]

    Mokady, R.; Hertz, A.; Aberman, K.; Pritch, Y.; and Cohen-Or, D. 2023. Null-text inversion for editing real images using guided diffusion models. In Proceedings of the IEEE conference on computer vision and pattern recognition, 6038--6047

    2023

  12. [20]

    Q.; and Dhariwal, P

    Nichol, A. Q.; and Dhariwal, P. 2021. Improved denoising diffusion probabilistic models. In International conference on machine learning, 8162--8171. PMLR

    2021

  13. [21]

    \"O zdenizci, O.; and Legenstein, R. 2023. Restoring vision in adverse weather conditions with patch-based denoising diffusion models. IEEE Transactions on Pattern Analysis and Machine Intelligence

    2023

  14. [22]

    Preechakul, K.; Chatthee, N.; Wizadwongsa, S.; and Suwajanakorn, S. 2022. Diffusion autoencoders: Toward a meaningful and decodable representation. In Proceedings of the IEEE conference on computer vision and pattern recognition, 10619--10629

    2022

  15. [23]

    T.; Popescu, M.; Khan, F

    Ristea, N.-C.; Croitoru, F.-A.; Ionescu, R. T.; Popescu, M.; Khan, F. S.; Shah, M.; et al. 2024. Self-distilled masked auto-encoders are efficient video anomaly detectors. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 15984--15995

    2024

  16. [24]

    Rombach, R.; Blattmann, A.; Lorenz, D.; Esser, P.; and Ommer, B. 2022. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE conference on computer vision and pattern recognition, 10684--10695

    2022

  17. [25]

    Ronneberger, O.; Fischer, P.; and Brox, T. 2015. U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention, 234--241. Springer

    2015

  18. [26]

    Roth, K.; Pemula, L.; Zepeda, J.; Sch \"o lkopf, B.; Brox, T.; and Gehler, P. 2022. Towards total recall in industrial anomaly detection. In Proceedings of the IEEE conference on computer vision and pattern recognition, 14318--14328

    2022

  19. [27]

    L.; Ghasemipour, K.; Gontijo Lopes, R.; Karagol Ayan, B.; Salimans, T.; et al

    Saharia, C.; Chan, W.; Saxena, S.; Li, L.; Whang, J.; Denton, E. L.; Ghasemipour, K.; Gontijo Lopes, R.; Karagol Ayan, B.; Salimans, T.; et al. 2022 a . Photorealistic text-to-image diffusion models with deep language understanding. In Advances in neural information processing...

    2022

  20. [28]

    J.; and Norouzi, M

    Saharia, C.; Ho, J.; Chan, W.; Salimans, T.; Fleet, D. J.; and Norouzi, M. 2022 b . Image super-resolution via iterative refinement. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(4): 4713--4726

    2022

  21. [29]

    P.; and Yang, W

    Song, Z.; Luo, R.; Yu, J.; Chen, Y.-P. P.; and Yang, W. 2023. Compact transformer tracker with correlative masked modeling. In Proceedings of the AAAI conference on artificial intelligence, volume 37, 2321--2329

    2023

  22. [30]

    P.; and Yang, W

    Song, Z.; Tang, Y.; Luo, R.; Ma, L.; Yu, J.; Chen, Y.-P. P.; and Yang, W. 2024. Autogenic language embedding for coherent point tracking. In Proceedings of the 32nd ACM International Conference on Multimedia, 2021--2030

    2024

  23. [31]

    P.; and Yang, W

    Song, Z.; Yu, J.; Chen, Y.-P. P.; and Yang, W. 2022. Transformer tracking with cyclic shifting window attention. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 8791--8800

    2022

  24. [32]

    Stephen, K.; and Menon, V. 2020. Learning Memory-guided Normality for Anomaly Detection. In Proceedings of the IEEE conference on computer vision and pattern recognition, 14360--14369

    2020

  25. [33]

    Teed, Z.; and Deng, J. 2020. Raft: Recurrent all-pairs field transforms for optical flow. In European Conference on Computer Vision, 402--419. Springer

    2020

  26. [34]

    Wolleb, J.; Bieder, F.; Sandk \"u hler, R.; and Cattin, P. C. 2022. Diffusion models for medical anomaly detection. In International Conference on Medical image computing and computer-assisted intervention, 35--45. Springer

    2022

  27. [35]

    Wu, J.; Fang, H.; Zhang, Y.; Yang, Y.; and Xu, Y. 2022. Medsegdiff: Medical image segmentation with diffusion probabilistic model. arXiv preprint arXiv:2211.00611

    2022 arXiv

  28. [36]

    Wu, P.; Liu, J.; Shi, Y.; Shao, F.; Wu, Z.; and Yang, Z. 2020. Not only Look, But Also Listen: Learning Multimodal Violence Detection Under Weak Supervision. In European Conference on Computer Vision, 322--339

    2020

  29. [37]

    Yan, C.; Zhang, S.; Liu, Y.; Pang, G.; and Wang, W. 2023. Feature Prediction Diffusion Model for Video Anomaly Detection. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 5527--5537

    2023

  30. [38]

    Yang, Z.; Liu, J.; Wu, Z.; Wu, P.; and Liu, X. 2023. Video Event Restoration Based on Keyframes for Video Anomaly Detection. In Proceedings of the IEEE conference on computer vision and pattern recognition, 14592--14601

    2023

  31. [39]

    Yang, Z.; Wu, P.; Liu, J.; and Liu, X. 2022. Dynamic local aggregation network with adaptive clusterer for anomaly detection. In European Conference on Computer Vision, 404--421. Springer

    2022

  32. [40]

    Ye, M.; Peng, X.; Gan, W.; Wu, W.; and Qiao, Y. 2019. Anopcn: Video anomaly detection via deep predictive coding network. In Proceedings of the 27th ACM international conference on multimedia, 1805--1813

    2019

  33. [41]

    Ye, Y.; Cai, J.; Zhou, H.; Li, G.; Zhang, Y.; Song, Z.; Gao, C.; Yu, J.; and Yang, W. 2024. Progressive Text-to-Image Diffusion with Soft Latent Direction. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 6693--6701

    2024

  34. [42]

    Yu, G.; Wang, S.; Cai, Z.; Zhu, E.; Xu, C.; Yin, J.; and Kloft, M. 2020. Cloze test helps: Effective video anomaly detection via learning to complete video events. In Proceedings of the 28th ACM international conference on multimedia, 583--591

    2020

  35. [43]

    Zhang, H.; Wang, Z.; Wu, Z.; and Jiang, Y.-G. 2023. DiffusionAD: Denoising Diffusion for Anomaly Detection. arXiv preprint arXiv:2303.08730

    2023 arXiv

  36. [44]

    Zhang, L.; and Agrawala, M. 2023. Adding conditional control to text-to-image diffusion models. arXiv preprint arXiv:2302.05543

    2023 arXiv

  37. [45]

    Zhao, M.; Liu, Y.; Liu, J.; Li, D.; and Zeng, X. 2022. LGN-Net: Local-Global Normality Network for Video Anomaly Detection. arXiv preprint arXiv:2211.07454

    2022 arXiv

  38. [46]

    Zhong, Y.; Chen, X.; Jiang, J.; and Ren, F. 2022. A cascade reconstruction model with generalization ability evaluation for anomaly detection in videos. Pattern Recognition, 122: 108336

    2022

  39. [47]

    Zhou, H.; Yu, J.; and Yang, W. 2023. Dual memory units with uncertainty regulation for weakly supervised video anomaly detection. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, 3769--3777

    2023

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.