Pith. sign in

REVIEW 4 major objections 4 minor 1 cited by

Frequency-Guided Diffusion Model with Perturbation Training for Skeleton-Based Video Anomaly Detection

T0 review · 4 major / 4 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 A frequency-guided diffusion model with perturbation training reports state-of-the-art accuracy on five skeleton-based video anomaly detection benchmarks by reconstructing global motion while preserving observed high-frequency detail.

desk verdict Solid empirical method paper for skeleton-based VAD with a useful perturbation-training + DCT-guidance combination, but the proof of Theorem IV.1 is invalid and the frequency-copy mechanism may miss high-frequency anomalies. read the letter →

arxiv 2412.03044 v2 pith:LALRBFY2 submitted 2024-12-04 cs.CV

Xiaofeng Tan , Hongsong Wang , Xin Geng , Liang Wang This is my paper · ORCID
classification cs.CV
keywords videoanomalydetectionskeleton-baseddiffusionmodelperturbationtrainingdiscretecosinetransformfrequency-guideddenoisingopen-setrobustnessreconstruction-based
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 show that two weaknesses of reconstruction-based video anomaly detection, poor robustness to unseen normal motions and unreliable reconstruction of local motion details, can both be fixed inside a single diffusion framework. The first fix is adversarial: a small generator learns perturbations that look like normal motions but maximize the diffusion model's reconstruction error, and the denoiser is trained against them to widen the set of motions it can reconstruct. The second fix is spectral: a 2D discrete cosine transform separates each motion into low-frequency (global) and high-frequency (local) parts, and at inference the observed high-frequency coefficients are copied straight into the reconstruction while only the low-frequency part is regenerated. The paper reports that this combination exceeds previous state-of-the-art AUC on Avenue, HR-Avenue, HR-STC, UBnormal, and HR-UBnormal. A sympathetic reader would care because it offers a concrete recipe for making open-set anomaly detection more forgiving of normal variation without needing anomalous training data.

What carries the argument

The load-bearing mechanism is the frequency-guided denoising loop, in which a 2D Discrete Cosine Transform (DCT) maps each motion into coefficients ordered by frequency, and a DCT-Mask splits them into a low-frequency set (the largest-magnitude coefficients, thresholded by $\lambda_{dct}$) and the complementary high-frequency set. At every denoising step the observed motion's high-frequency coefficients are copied into the fused motion while the denoiser's low-frequency coefficients are kept; the inverse DCT then returns a motion that the denoiser refines toward the normal-motion distribution. The second mechanism is the perturbation generator $G_\phi$, a lightweight network producing sign-constrained perturbations $\lambda_p \cdot \mathrm{sign}(G_\phi(x))$ that maximize the denoising loss, trained adversarially against the noise predictor $\epsilon_\theta$ so that the reconstruction domain expands toward unseen normal motions. A theorem in the appendix asserts that the generated perturbed motion stays within a bounded neighborhood of the observed motion while increasing reconstruction error.

What would settle it

Build or collect skeleton sequences where the global motion is a normal walk but a single joint (for example, the wrist) performs a small rapid tremor that is abnormal in context. If the proposed method's reconstruction error for these sequences stays as low as for genuinely normal sequences, the frequency-fusion assumption fails; a simpler check is to run the released code on such sequences and compare AUC against a baseline that uses full reconstruction error.

Watch

Extended reading notes

Core claim

The central discovery claimed is that anomaly detection accuracy improves when a diffusion model is deliberately trained on adversarially chosen near-normal perturbations and then, at inference, is allowed to keep the high-frequency DCT coefficients of the observed motion instead of reconstructing them. The paper argues that high-frequency motion components (stride length, arm-swing amplitude, joint-level habits) vary from person to person and are therefore not reliable signals of abnormality, while low-frequency components encode the global structure that determines whether a motion is normal or anomalous. Under the fusion rule $y^c_t = y^o_t \odot M_h(y^o_t) + y^g_t \odot M_l(y^g_t)$, the model regenerates only low-frequency content and preserves observed high-frequency detail, which the paper shows yields higher AUC than reconstructing all frequencies equally. Combined with perturbation training, the method reaches 88.0 AUC on Avenue, 90.7 on HR-Avenue, 78.6 on HR-STC, 68.9 on UBnormal, and 69.0 on HR-UBnormal, outperforming the listed state-of-the-art methods including prediction-based and hybrid ones. The paper also shows that perturbation training keeps performance stable when inference-time perturbation intensity increases, whereas a model trained without it degrades sharply.

Load-bearing premise

The method assumes that abnormal motions always change the global, low-frequency structure of a skeleton sequence, while high-frequency components are only person-specific style, so an anomaly that lives entirely in small, local joint details would be copied into the reconstruction and missed.

Editorial extensions

If this is right

  • Unseen normal motions with stylistic variations should no longer be systematically misclassified, because perturbation training deliberately exposes the model to near-normal inputs it would otherwise reconstruct poorly.
  • Reconstruction error becomes a more reliable anomaly score, because the score no longer depends on the model's ability to reproduce person-specific high-frequency details.
  • The approach remains competitive with supervised and weakly supervised methods (for example, 68.9 AUC on UBnormal) while using only 556K parameters and no anomaly labels.
  • Ablations indicate that both components contribute: removing perturbation training drops AUC by up to 4.6 points, and replacing the DCT-Mask with a temporal mask drops it by up to 0.9 points.

Reading between the lines

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

  • The frequency-splitting idea is not tied to skeletons; the same DCT-fusion rule could be tested on RGB video features or general time-series anomaly detection, where global trend versus local detail separation may behave differently.
  • The paper's assumption that anomalies live in low-frequency structure is testable by constructing anomalies confined to a single joint or small joint group; if such anomalies pass through with low error, the fusion rule needs an exception mechanism.
  • The threshold $\lambda_{dct}$ is tuned per dataset, suggesting the optimal low/high split depends on the motion distribution; an adaptive threshold derived from training-set statistics could remove this per-dataset tuning and is a direct extension of the paper's own sensitivity analysis.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper proposes FG-Diff, a skeleton-based video anomaly detection method combining a diffusion model trained with adversarial perturbation samples (a learned perturbation generator) and a frequency-guided denoising process. At training time, a perturbation generator maximizes the denoising loss within a bounded neighborhood, and the denoiser minimizes it on the perturbed samples. At inference, the method applies 2D-DCT masks to fuse high-frequency components of the observed motion with low-frequency components of the generated motion at each denoising step, then scores anomalies by reconstruction error. The authors report state-of-the-art AUC on Avenue, HR-Avenue, HR-STC, UBnormal, and HR-UBnormal, together with ablations and a robustness analysis.

Significance. If the empirical results hold, the paper would contribute a practical recipe for improving open-set robustness of skeleton-based reconstruction methods, and the frequency-separation idea is an interesting departure from uniform reconstruction targets. The paper is commendable for evaluating on five datasets, including open-set UR-normal, for providing a robustness analysis of perturbation training, and for reporting parameter efficiency against supervised baselines. However, the central theoretical justification (Theorem IV.1) is invalid, and the frequency-copy fusion rule has an untested failure mode that is load-bearing for the anomaly-detection claim. The per-dataset tuning of the frequency threshold and the lack of statistical significance reporting further weaken the strength of the empirical claims. The core idea is defensible, but the manuscript needs substantial revision before the claims can be accepted.

major comments (4)
  1. [Section VII-A / Theorem IV.1 (Eqs. 25-27)] The above comment is a major comment with invalid proof. Please provide a corrected version.
  2. [Section IV-C, Eq. (18), and Algorithm 2] The above comment is a major comment. Please provide a corrected version.
  3. [Algorithm 2, lines 6-7, and Table IV] The above comment is a major comment. Please provide a corrected version.
  4. [Tables I-III] The above comment is a major comment. Please provide a corrected version.
minor comments (4)
  1. [Eq. (22)] The above comment is a minor comment. Please provide a corrected version.
  2. [Algorithm 2, line 6] The above comment is a minor comment. Please provide a corrected version.
  3. [Section V-C, Table III] The above comment is a minor comment. Please provide a corrected version.
  4. [Fig. 6] The above comment is a minor comment. Please provide a corrected version.

Circularity Check

1 steps flagged · score 2.0 of 10

Central empirical claim is independently benchmarked; only Theorem IV.1 restates its training objective by definition.

  1. self definitional [Section IV-B, Theorem IV.1, and Appendix VII-A, Eqs. (24)-(27)]
    "Theorem IV.1 ... (b) Increased reconstruction error: S(xo) − S(ˆxo) ≤ 0. ... Eq. (4) demonstrates that the anomaly score S(x) is directly measured by reconstruction error, i.e., S(x) = L(x, θ). ... Gϕ is optimized by: max_ϕ L(xo + λpsign(Gϕ(xo)), θ). ... The following relationship holds: ˆxo = arg max_{ˆxo} L(ˆxo, θ), ... ⇐⇒ ˆxo = arg max_{ˆxo} S(ˆxo), ... Thus, we have: S(ˆxo) ≥ S(xo) ⇐⇒ S(xo) − S(ˆxo) ≤ 0."

    The theorem's claimed conclusion (b) is a restatement of the training objective: because the paper defines S(x) = L(x, θ), 'maximize L over the perturbed motion' is by definition 'maximize reconstruction error over the perturbed motion.' The proof's Eq. (26)-(27) simply rewrites the argmax as an inequality, so the 'increased reconstruction error' is assumed by the max over ϕ rather than derived from the training dynamics or from generalization. This is a definitional tautology, not an independent result. The circularity is minor because perturbation training's benefit is separately supported by the Table IV robustness experiments, so the theorem is not the sole evidence for the central empirical claim.

full rationale

The main accuracy claims are evaluated on held-out test sets of Avenue, HR-Avenue, HR-STC, UBnormal, and HR-UBnormal against external baselines, so the central empirical result is not circular. The DCT fusion in Eq. (18) is a design choice: it states that the method scores anomalies through low-frequency reconstruction error while copying observed high-frequency coefficients as guidance. This is an inductive bias and a possible failure mode for high-frequency-only anomalies, but it is not a circular derivation because the paper does not claim to detect such anomalies from the copied coefficients. The only genuine circular step is Theorem IV.1, where the proof that perturbation increases reconstruction error reduces to the definition S = L and to the generator's maximization objective. Since this theorem is formal support rather than the empirical backbone, and since the robustness benefit is independently demonstrated in Table IV across varying inference perturbation intensities, the overall circularity score is low.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The method's performance rests on two tuned hyperparameters (λ_p, λ_dct) and two domain assumptions about motion frequency semantics and perturbation validity. The theoretical guarantee for perturbation training (Theorem IV.1) is not actually established, because the proof assumes the generator is the argmax of the loss over the perturbation set. The empirical evaluation is externally benchmarked on test sets, so the core claim is not circular, but the theoretical scaffolding is weak.

free parameters (3)
  • Perturbation magnitude λ_p = 0.1 (all datasets)
    Controls the L∞ bound on adversarial perturbations in training (Eq. 10) and inference (Algorithm 2). Chosen by hand; the robustness table suggests performance is stable around this value.
  • DCT mask threshold λ_dct = 0.9 for UBnormal/HR-UBnormal, 0.1 for others
    Percentage of top-magnitude DCT coefficients retained as low-frequency. Tuned per dataset; sensitivity curves in Fig. 6 show AUC varies by up to 3.6 points.
  • Condition-code DCT top-k size k = not reported
    The condition code uses the top k DCT coefficients (DCT_k), but k is not specified in the paper. This is an additional hyperparameter affecting the information provided to the denoiser.
assumptions (4)
  • domain assumption Anomalies are primarily encoded in low-frequency (global) motion structure; high-frequency components are person-specific style and not reliable indicators of abnormality.
    This premise motivates the fusion rule in Eq. 18, which copies observed high-frequency coefficients into the reconstruction. It is stated in Section IV-C but not validated with experiments isolating high-frequency anomalies.
  • domain assumption The perturbation generator, after adversarial training, produces perturbed motions that lie near the observed normal motion and increase the reconstruction model's loss.
    Theorem IV.1 is intended to guarantee this, but the proof is invalid (Appendix VII-A). The method's robustness improvement depends on this behavior holding on unseen normal motions.
  • domain assumption A diffusion model trained on normal motions assigns lower reconstruction error to normal motions than to anomalous motions after the frequency-guided fusion.
    This is the standard VAD assumption inherited from MoCoDAD [8]; the paper does not prove it but relies on it for the anomaly score in Eq. 4.
  • standard math The 2D DCT applied to a reshaped skeleton sequence meaningfully separates temporal and spatial frequency content.
    DCT is a standard transform; the specific reshaping of N frames by C*J features is a modeling choice, but the math is standard.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Frequency-Guided Diffusion Model with Perturbation Training for Skeleton-Based Video Anomaly Detection." pith.science (2026). https://pith.science/paper/LALRBFY2

@misc{pith2026241203044,
  author       = {Pith},
  title        = {Pith review of: Frequency-Guided Diffusion Model with Perturbation Training for Skeleton-Based Video Anomaly Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LALRBFY2}},
  note         = {Machine review of arXiv:2412.03044}
}
read the original abstract

Video anomaly detection (VAD) is a vital yet complex open-set task in computer vision, commonly tackled through reconstruction-based methods. However, these methods struggle with two key limitations: (1) insufficient robustness in open-set scenarios, where unseen normal motions are frequently misclassified as anomalies, and (2) an overemphasis on, but restricted capacity for, local motion reconstruction, which are inherently difficult to capture accurately due to their diversity. To overcome these challenges, we introduce a novel frequency-guided diffusion model with perturbation training. First, we enhance robustness by training a generator to produce perturbed samples, which are similar to normal samples and target the weakness of the reconstruction model. This training paradigm expands the reconstruction domain of the model, improving its generalization to unseen normal motions. Second, to address the overemphasis on motion details, we employ the 2D Discrete Cosine Transform (DCT) to separate high-frequency (local) and low-frequency (global) motion components. By guiding the diffusion model with observed high-frequency information, we prioritize the reconstruction of low-frequency components, enabling more accurate and robust anomaly detection. Extensive experiments on five widely used VAD datasets demonstrate that our approach surpasses state-of-the-art methods, underscoring its effectiveness in open-set scenarios and diverse motion contexts. Our project website is https://xiaofeng-tan.github.io/projects/FG-Diff/index.html.

Figures

Figures reproduced from arXiv: 2412.03044 by the authors.

Figure 1
Figure 1. The data illustration. (a) The training and testing data, where the [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Comparison between our proposed method (green) and existing [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. The framework of the proposed method. The model is trained utilizing generated perturbation examples. The training phase includes two processes: [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: The illustration of perturbation training. In Fig. (a), the green and [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: The visualization of human motions processed by 2D-DCT. (a) original [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 7
Figure 7. Figure 7: Anomaly score curves on the Avenue and HR-UBnormal datasets. (a) [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]

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. The Evolution of Video Anomaly Detection: A Unified Framework from DNN to MLLM

    cs.CV 2025-07 conditional novelty 4.0 of 10

    The paper organizes VAD methods into a five-dimension framework spanning task objective, modality, input, architecture, and optimization, with emphasis on MLLM/LLM-era work.

Reference graph

Works this paper leans on

57 extracted references · 52 canonical work pages · cited by 1 Pith paper

  1. [1]

    Appearance-motion memory consistency network for video anomaly detection,

    R. Cai, H. Zhang, W. Liu, S. Gao, and Z. Hao, “Appearance-motion memory consistency network for video anomaly detection,” in Proceed- ings of the AAAI Conference on Artificial Intelligence , vol. 35, no. 2, 2021, pp. 938–946

    work page 2021

  2. [2]

    Video anomaly detection with sparse coding inspired deep neural networks,

    W. Luo, W. Liu, D. Lian, J. Tang, L. Duan, X. Peng, and S. Gao, “Video anomaly detection with sparse coding inspired deep neural networks,” IEEE Transactions on Pattern Analysis and Machine Intelli- gence, vol. 43, no. 3, pp. 1070–1084, 2021

    work page 2021

  3. [3]

    Gaussian process regression-based video anomaly detection and localization with hierar- chical feature representation,

    K.-W. Cheng, Y .-T. Chen, and W.-H. Fang, “Gaussian process regression-based video anomaly detection and localization with hierar- chical feature representation,” IEEE Transactions on Image Processing , vol. 24, no. 12, pp. 5288–5301, 2015

    work page 2015

  4. [4]

    Video anomaly detection with compact feature sets for online performance,

    R. Leyva, V . Sanchez, and C.-T. Li, “Video anomaly detection with compact feature sets for online performance,” IEEE Transactions on Image Processing, vol. 26, no. 7, pp. 3463–3478, 2017

    work page 2017

  5. [5]

    Detection of dynamic background due to swaying movements from motion features,

    D.-S. Pham, O. Arandjelovi ´c, and S. Venkatesh, “Detection of dynamic background due to swaying movements from motion features,” IEEE Transactions on Image Processing , vol. 24, no. 1, pp. 332–344, 2015

    work page 2015

  6. [6]

    Ubnormal: New benchmark for supervised open-set video anomaly detection,

    A. Acsintoae, A. Florescu, M.-I. Georgescu, T. Mare, P. Sumedrea, R. T. Ionescu, F. S. Khan, and M. Shah, “Ubnormal: New benchmark for supervised open-set video anomaly detection,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022, pp. 20 111–20 121

    work page 2022

  7. [7]

    Learning regularity in skeleton trajectories for anomaly detection in videos,

    R. Morais, V . Le, T. Tran, B. Saha, M. Mansour, and S. Venkatesh, “Learning regularity in skeleton trajectories for anomaly detection in videos,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019, pp. 11 988–11 996

    work page 2019

  8. [8]

    Multimodal motion conditioned diffusion model for skeleton-based video anomaly detection,

    A. Flaborea, L. Collorone, G. M. D’Amely Di Melendugno, S. D’Arrigo, B. Prenkaj, and F. Galasso, “Multimodal motion conditioned diffusion model for skeleton-based video anomaly detection,” in IEEE/CVF In- ternational Conference on Computer Vision , 2023, pp. 10 284–10 295

    work page 2023

Show all 57 references
  1. [9]

    Video anomaly detection via spatio- temporal pseudo-anomaly generation: A unified approach,

    A. K. Rai, T. Krishna, F. Hu, A. Drimbarean, K. McGuinness, A. F. Smeaton, and N. E. O’connor, “Video anomaly detection via spatio- temporal pseudo-anomaly generation: A unified approach,” in Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, ...

    2024

  2. [10]

    Unbiased multiple instance learning for weakly supervised video anomaly detection,

    H. Lv, Z. Yue, Q. Sun, B. Luo, Z. Cui, and H. Zhang, “Unbiased multiple instance learning for weakly supervised video anomaly detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 8022–8031

    2023

  3. [11]

    Hierarchical semantic contrast for scene-aware video anomaly detection,

    S. Sun and X. Gong, “Hierarchical semantic contrast for scene-aware video anomaly detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 22 846–22 856

    2023

  4. [12]

    A new comprehensive bench- mark for semi-supervised video anomaly detection and anticipation,

    C. Cao, Y . Lu, P. Wang, and Y . Zhang, “A new comprehensive bench- mark for semi-supervised video anomaly detection and anticipation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 20 392–20 401

    2023

  5. [13]

    Video event restoration based on keyframes for video anomaly detection,

    Z. Yang, J. Liu, Z. Wu, P. Wu, and X. Liu, “Video event restoration based on keyframes for video anomaly detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 14 592–14 601

    2023

  6. [14]

    Normalizing flows for human pose anomaly detection,

    O. Hirschorn and S. Avidan, “Normalizing flows for human pose anomaly detection,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 13 545–13 554

    2023

  7. [15]

    Skeletal video anomaly detection using deep learning: Survey, challenges, and future directions,

    P. K. Mishra, A. Mihailidis, and S. S. Khan, “Skeletal video anomaly detection using deep learning: Survey, challenges, and future directions,” IEEE Transactions on Emerging Topics in Computational Intelligence , vol. 8, no. 2, pp. 1073–1085, 2024

    2024

  8. [16]

    Deep 3d human pose estimation: A review,

    J. Wang, S. Tan, X. Zhen, S. Xu, F. Zheng, Z. He, and L. Shao, “Deep 3d human pose estimation: A review,” Computer Vision and Image Understanding, vol. 210, p. 103225, 2021

    2021

  9. [17]

    Holistic representation learning for multitask trajectory anomaly detection,

    A. Stergiou, B. De Weerdt, and N. Deligiannis, “Holistic representation learning for multitask trajectory anomaly detection,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision , 2024, pp. 6729–6739

    2024

  10. [18]

    Multi- timescale trajectory prediction for abnormal human activity detection,

    R. Rodrigues, N. Bhargava, R. Velmurugan, and S. Chaudhuri, “Multi- timescale trajectory prediction for abnormal human activity detection,” in IEEE Winter Conference on Applications of Computer Vision , 2020, pp. 2615–2623

    2020

  11. [19]

    A kalman variational autoencoder model assisted by odometric clustering for video frame prediction and anomaly detection,

    G. Slavic, A. S. Alemaw, L. Marcenaro, D. Mart ´ın G ´omez, and C. Regazzoni, “A kalman variational autoencoder model assisted by odometric clustering for video frame prediction and anomaly detection,” IEEE Transactions on Image Processing , vol. 32, pp. 415–429, 2023

    2023

  12. [20]

    Easynet: An easy network for 3d industrial anomaly detection,

    R. Chen, G. Xie, J. Liu, J. Wang, Z. Luo, J. Wang, and F. Zheng, “Easynet: An easy network for 3d industrial anomaly detection,” in Proceedings of the 31st ACM International Conference on Multimedia , IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 00, NO. 0, MARCH 2025 11 ser. MM...

    2025

  13. [21]

    Look inside for more: Internal spatial modality perception for 3d anomaly detection,

    H. Liang, G. Xie, C. Hou, B. Wang, C. Gao, and J. Wang, “Look inside for more: Internal spatial modality perception for 3d anomaly detection,” 2025

    2025

  14. [22]

    Time series anomaly detection with adversarial reconstruction net- works,

    S. Liu, B. Zhou, Q. Ding, B. Hooi, Z. Zhang, H. Shen, and X. Cheng, “Time series anomaly detection with adversarial reconstruction net- works,” IEEE Transactions on Knowledge and Data Engineering , vol. 35, no. 4, pp. 4293–4306, 2022

    2022

  15. [23]

    Toward video anomaly retrieval from video anomaly detection: New benchmarks and model,

    P. Wu, J. Liu, X. He, Y . Peng, P. Wang, and Y . Zhang, “Toward video anomaly retrieval from video anomaly detection: New benchmarks and model,” IEEE Transactions on Image Processing , vol. 33, pp. 2213– 2225, 2024

    2024

  16. [24]

    Omni-frequency channel-selection representations for unsupervised anomaly detection,

    Y . Liang, J. Zhang, S. Zhao, R. Wu, Y . Liu, and S. Pan, “Omni-frequency channel-selection representations for unsupervised anomaly detection,” IEEE Transactions on Image Processing, vol. 32, pp. 4327–4340, 2023

    2023

  17. [25]

    Advancing video anomaly detection: A bi-directional hybrid framework for enhanced single- and multi-task approaches,

    G. Shen, Y . Ouyang, J. Lu, Y . Yang, and V . Sanchez, “Advancing video anomaly detection: A bi-directional hybrid framework for enhanced single- and multi-task approaches,” IEEE Transactions on Image Pro- cessing, vol. 33, pp. 6865–6880, 2024

    2024

  18. [26]

    Variational ab- normal behavior detection with motion consistency,

    J. Li, Q. Huang, Y . Du, X. Zhen, S. Chen, and L. Shao, “Variational ab- normal behavior detection with motion consistency,” IEEE Transactions on Image Processing , vol. 31, pp. 275–286, 2022

    2022

  19. [27]

    Remembering history with convolutional lstm for anomaly detection,

    W. Luo, W. Liu, and S. Gao, “Remembering history with convolutional lstm for anomaly detection,” in 2017 IEEE International Conference on Multimedia and Expo (ICME) , 2017, pp. 439–444

    2017

  20. [28]

    Synthetic temporal anomaly guided end-to-end video anomaly detection,

    M. Astrid, M. Z. Zaheer, and S.-I. Lee, “Synthetic temporal anomaly guided end-to-end video anomaly detection,” in IEEE/CVF International Conference on Computer Vision Workshops , 2021, pp. 207–214

    2021

  21. [29]

    Learning not to reconstruct anomalies,

    M. Astrid, M. Z. Zaheer, J.-Y . Lee, and S.-I. Lee, “Learning not to reconstruct anomalies,” in British Machine Vision Conference , 2021

    2021

  22. [30]

    Learning temporal regularity in video sequences,

    M. Hasan, J. Choi, J. Neumann, A. K. Roy-Chowdhury, and L. S. Davis, “Learning temporal regularity in video sequences,” in IEEE Conference on Computer Vision and Pattern Recognition , 2016, pp. 733–742

    2016

  23. [31]

    A revisit of sparse coding based anomaly detection in stacked rnn framework,

    W. Luo, W. Liu, and S. Gao, “A revisit of sparse coding based anomaly detection in stacked rnn framework,” in IEEE International Conference on Computer Vision , 2017, pp. 341–349

    2017

  24. [32]

    Context recovery and knowledge retrieval: A novel two-stream framework for video anomaly detection,

    C. Cao, Y . Lu, and Y . Zhang, “Context recovery and knowledge retrieval: A novel two-stream framework for video anomaly detection,” IEEE Transactions on Image Processing , vol. 33, pp. 1810–1825, 2024

    2024

  25. [33]

    Memorizing normality to detect anomaly: Memory-augmented deep autoencoder for unsupervised anomaly detec- tion,

    D. Gong, L. Liu, V . Le, B. Saha, M. R. Mansour, S. Venkatesh, and A. Van Den Hengel, “Memorizing normality to detect anomaly: Memory-augmented deep autoencoder for unsupervised anomaly detec- tion,” in IEEE/CVF International Conference on Computer Vision, 2019, pp. 1705–1714

    2019

  26. [34]

    Learning memory-guided normality for anomaly detection,

    H. Park, J. Noh, and B. Ham, “Learning memory-guided normality for anomaly detection,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, pp. 14 360–14 369

    2020

  27. [35]

    A hybrid video anomaly detection framework via memory-augmented flow reconstruction and flow-guided frame prediction,

    Z. Liu, Y . Nie, C. Long, Q. Zhang, and G. Li, “A hybrid video anomaly detection framework via memory-augmented flow reconstruction and flow-guided frame prediction,” in IEEE/CVF International Conference on Computer Vision , 2021, pp. 13 568–13 577

    2021

  28. [36]

    Usdrl: Unified skeleton-based dense representation learning with multi-grained feature decorrelation,

    W. Weng, H. Wang, J. He, L. He, and G. Xie, “Usdrl: Unified skeleton-based dense representation learning with multi-grained feature decorrelation,” arXiv preprint arXiv:2412.09220 , 2024

    2024 arXiv

  29. [37]

    Sopo: Text-to-motion generation using semi-online preference optimization,

    X. Tan, H. Wang, X. Geng, and P. Zhou, “Sopo: Text-to-motion generation using semi-online preference optimization,” arXiv preprint arXiv:2412.05095, 2024

    2024

  30. [38]

    Graph embedded pose clustering for anomaly detection,

    A. Markovitz, G. Sharir, I. Friedman, L. Zelnik-Manor, and S. Avidan, “Graph embedded pose clustering for anomaly detection,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020, pp. 10 536–10 544

    2020

  31. [39]

    Contracting skeletal kinematics for human-related video anomaly detection,

    A. Flaborea, G. M. D. di Melendugno, S. D’arrigo, M. A. Sterpa, A. Sampieri, and F. Galasso, “Contracting skeletal kinematics for human-related video anomaly detection,” Pattern Recognition , p. 110817, 2024

    2024

  32. [40]

    A hierarchical spatio-temporal graph convolutional neural network for anomaly detec- tion in videos,

    X. Zeng, Y . Jiang, W. Ding, H. Li, Y . Hao, and Z. Qiu, “A hierarchical spatio-temporal graph convolutional neural network for anomaly detec- tion in videos,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 33, no. 1, pp. 200–212, 2023

    2023

  33. [41]

    Hi- erarchical graph embedded pose regularity learning via spatio-temporal transformer for abnormal behavior detection,

    C. Huang, Y . Liu, Z. Zhang, C. Liu, J. Wen, Y . Xu, and Y . Wang, “Hi- erarchical graph embedded pose regularity learning via spatio-temporal transformer for abnormal behavior detection,” in Proceedings of the 30th ACM International Conference on Multimedia, ser. MM ’22. New ...

    2022

  34. [42]

    Regularity learning via explicit distribution modeling for skeletal video anomaly detection,

    S. Yu, Z. Zhao, H. Fang, A. Deng, H. Su, D. Wang, W. Gan, C. Lu, and W. Wu, “Regularity learning via explicit distribution modeling for skeletal video anomaly detection,” IEEE Transactions on Circuits and Systems for Video Technology, 2023

    2023

  35. [43]

    Three-way decision-based co- detection for outliers,

    X. Tan, C. Gao, J. Zhou, and J. Wen, “Three-way decision-based co- detection for outliers,” International Journal of Approximate Reasoning, vol. 160, p. 108971, 2023

    2023

  36. [44]

    Fuzzy granule density-based outlier detection with multi-scale granular balls,

    C. Gao, X. Tan, J. Zhou, W. Ding, and W. Pedrycz, “Fuzzy granule density-based outlier detection with multi-scale granular balls,” IEEE Transactions on Knowledge and Data Engineering , vol. 37, no. 3, pp. 1182–1197, 2025

    2025

  37. [45]

    Explaining and harnessing adversarial examples,

    I. J. Goodfellow, J. Shlens, and C. Szegedy, “Explaining and harnessing adversarial examples,” in International Conference on Learning Repre- sentations, 2015

    2015

  38. [46]

    Enhancing the reliability of out- of-distribution image detection in neural networks,

    S. Liang, Y . Li, and R. Srikant, “Enhancing the reliability of out- of-distribution image detection in neural networks,” in International Conference on Learning Representations , 2018

    2018

  39. [47]

    Generalized odin: Detecting out-of-distribution image without learning from out-of-distribution data,

    Y .-C. Hsu, Y . Shen, H. Jin, and Z. Kira, “Generalized odin: Detecting out-of-distribution image without learning from out-of-distribution data,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020, pp. 10 948–10 957

    2020

  40. [48]

    A mathematical analysis of the dct coefficient distributions for images,

    E. Y . Lam and J. W. Goodman, “A mathematical analysis of the dct coefficient distributions for images,” IEEE Transactions on Image Processing, vol. 9, no. 10, pp. 1661–1666, 2000

    2000

  41. [49]

    Abnormal event detection at 150 fps in matlab,

    C. Lu, J. Shi, and J. Jia, “Abnormal event detection at 150 fps in matlab,” in IEEE International Conference on Computer Vision , 2013, pp. 2720– 2727

    2013

  42. [50]

    Posecvae: Anomalous human activity detection,

    Y . Jain, A. K. Sharma, R. Velmurugan, and B. Banerjee, “Posecvae: Anomalous human activity detection,” in International Conference on Pattern Recognition, 2021, pp. 2927–2934

    2021

  43. [51]

    Future frame prediction network for video anomaly detection,

    W. Luo, W. Liu, D. Lian, and S. Gao, “Future frame prediction network for video anomaly detection,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 44, no. 11, pp. 7505–7520, 2022

    2022

  44. [52]

    Eval: Explainable video anomaly localization,

    A. Singh, M. J. Jones, and E. G. Learned-Miller, “Eval: Explainable video anomaly localization,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 18 717–18 726

    2023

  45. [53]

    Open- vocabulary video anomaly detection,

    P. Wu, X. Zhou, G. Pang, Y . Sun, J. Liu, P. Wang, and Y . Zhang, “Open- vocabulary video anomaly detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 18 297–18 307

    2024

  46. [54]

    Real-world anomaly detection in surveillance videos,

    W. Sultani, C. Chen, and M. Shah, “Real-world anomaly detection in surveillance videos,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018, pp. 6479–6488

    2018

  47. [55]

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

    M.-I. Georgescu, A. B ˘arb˘al˘au, R. T. Ionescu, F. Shahbaz Khan, M. Popescu, and M. Shah, “Anomaly detection in video via self- supervised and multi-task learning,” in IEEE/CVF Conference on Com- puter Vision and Pattern Recognition , 2021, pp. 12 737–12 747

    2021

  48. [56]

    Is space-time attention all you need for video understanding?

    G. Bertasius, H. Wang, and L. Torresani, “Is space-time attention all you need for video understanding?” in Proceedings of the International Conference on Machine Learning , M. Meila and T. Zhang, Eds., vol

  49. [139]

    PMLR, 18–24 Jul 2021, pp. 813–824

    2021

Pith tools

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