Pith. sign in

REVIEW 3 major objections 5 minor 40 references

TeG: Temporal-Granularity Method for Anomaly Detection with Attention in Smart City Surveillance

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Pith's one-line read Fusing short-, medium-, and long-term video features with attention achieves 87.16% AUC on UCF-Crime and detects unseen anomalies in a city field lab.

desk verdict Competitive numbers and a real field deployment, but the core temporal-granularity claim is untested because there are no ablations. read the letter →

arxiv 2411.11003 v1 pith:7L5C2X3F submitted 2024-11-17 cs.CV

classification cs.CV
keywords videoanomalydetectiontemporalgranularitymulti-headcross-attentionself-attentionweaklysupervisedlearningsmartcitysurveillanceUCF-Crimereal-timedeployment
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 argues that surveillance anomalies have very different durations, and a detector that looks at only one time scale will miss either abrupt actions or slow-developing events. It introduces TeG, which splits each video segment into chunks of 8, 32, and 64 frames, extracts features for each scale with a Video Swin Transformer, and fuses the three feature sets with multi-head cross-attention and self-attention before classification. On public benchmarks the model reports 87.16% area under the curve on UCF-Crime, 95.32% on ShanghaiTech, and 84.57% average precision on XD-Violence. The paper also extends UCF-Crime with additional smart-city anomaly types and validates the system in a field lab, where it flags both trained (seen) and untrained (unseen) anomaly classes at roughly 80% accuracy. The point is that multi-scale temporal fusion can make weakly supervised anomaly detection work for the full range of real-world event durations.

What carries the argument

The central mechanism is the temporal-granularity feature fusion built from multi-head cross-attention (MCA) and multi-head self-attention (MSA). MCA lets feature sets from different time scales attend to each other, forming pairwise fused matrices $F_{SM}$, $F_{ML}$, and $F_{SL}$; MSA re-weights each scale by global correlations across all 32 video segments, forming $F_{SML}$. The four matrices are concatenated with a residual connection, so the classifier sees both cross-scale correlations and whole-clip temporal context. The same machinery is what the paper credits for detecting anomalies of different durations.

What would settle it

Train the identical Video Swin Transformer and loss on UCF-Crime using only the 64-frame (long) granularity, with no cross-scale attention, and measure AUC. If the score matches or exceeds 87.16%, the temporal-granularity and attention-fusion mechanism is not what carries the result.

Watch

Extended reading notes

Core claim

TeG's central claim is that explicitly modeling temporal granularity improves weakly supervised video anomaly detection. Each video is divided into 32 segments, and each segment is re-chunked at three granularities, 8, 32, and 64 frames, to capture short, medium, and long dynamics. The Video Swin Transformer produces a feature vector per chunk, averaged per segment into three representations $F_S$, $F_M$, and $F_L$. Multi-head cross-attention correlates pairs of scales and self-attention models dependencies across segments, producing four matrices that are concatenated with a residual connection and fed to a three-layer classifier trained with the top-k MIL ranking loss with feature-magnitude learning. The authors report that this design reaches 87.16% AUC on UCF-Crime, 95.32% AUC on ShanghaiTech, and 84.57% AP on XD-Violence, and that in a field lab with five cameras it detects both the anomaly classes it trained on and three untrained ones, with 79.81% accuracy and an F1 score of 0.88 over all 109 recorded anomalies.

Load-bearing premise

The results are credited to multi-scale temporal fusion, but the paper does not compare against the same backbone and loss without that fusion, so the key assumption is that the granularity design, not the pretrained video model or the training loss, is what produces the reported accuracy.

Editorial extensions

If this is right

  • A single model can flag both brief anomalies like thrown objects or accidents and longer ones like fights, because it pools evidence from 8-frame, 32-frame, and 64-frame chunks.
  • On the UCF-Crime and XD-Violence benchmarks, the multi-scale approach exceeds the reported scores of the same-loss baseline, suggesting the fusion adds detection power for complex real-world scenes.
  • A weakly supervised system trained only on video-level labels can generalize to anomaly classes it never saw, as the field-lab results on unseen classes show.
  • The system can run in near real time: processing one 10-second segment takes about 1.5 seconds on a GTX-2080Ti, so control-room operators can receive alerts with type, time, camera, and a GIF of the event.
  • Adding new anomaly classes to UCF-Crime (from 13 to 17) supports evaluation of smart-city scenarios such as dangerous throwing, littering, and traffic accidents.

Reading between the lines

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

  • The paper does not isolate the contribution of the attention fusion; a reader could train the same backbone and loss at a single granularity and compare. If a single scale matches the reported numbers, then the multi-scale attention design is not the decisive factor.
  • Granularity settings of 8, 32, and 64 frames are fixed empirically; sweeping these values would show how sensitive the result is and whether adaptive scales could improve it.
  • Extending the field-lab protocol into a standardized open-set benchmark would let other systems measure how well they generalize to anomaly classes absent from training.
  • Inspecting the attention weights could test the paper's underlying duration hypothesis directly: the model should concentrate on short scales for fast events and long scales for slow ones.
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

3 major / 5 minor

Summary. The paper proposes TeG, a weakly-supervised video anomaly detection method that extracts spatio-temporal features from a Video Swin Transformer backbone at three temporal granularities (8, 32, and 64 frames), fuses them with multi-head cross-attention and multi-head self-attention blocks plus a residual connection, and trains with the RTFM feature-magnitude and MIL ranking losses. The method is evaluated on UCF-Crime, ShanghaiTech, and XD-Violence, reporting 87.16% AUC, 95.32% AUC, and 84.57% AP, respectively. The paper also extends UCF-Crime with three Smart-City anomaly classes and reports a field-lab experiment on 109 staged samples, including a real-time deployment with anomaly alerts to control-room operators.

Significance. If the reported results are reproducible, TeG is competitive on two of the three standard benchmarks, and the deployment component is a useful industrial validation. The dataset extension is a modest but useful contribution, and the execution-time analysis is a strength. However, the central methodological claim — that multi-granularity temporal fusion via attention is the active ingredient behind the reported gains — is not supported by any controlled comparison in the current manuscript. The field-lab evidence is also too small and coarsely aggregated to substantiate generalization to unseen anomaly classes.

major comments (3)
  1. [Section IV.B, Table I] The comparison in Table I contrasts TeG with prior methods that differ simultaneously in backbone, fusion architecture, and loss function. Because TeG uses the same RTFM loss, a minimal control is RTFM with the same VST backbone and single-scale features, or TeG with the attention fusion removed. Without such ablations, the observed differences (87.16 vs 84.30 AUC on UCF-Crime, 84.57 vs 77.81 AP on XD-Violence) cannot be attributed to temporal-granularity fusion rather than to the VST backbone or the inherited loss. This directly affects the paper's central claim and should be addressed with controlled experiments.
  2. [Section IV.C, Table III] The field-lab validation uses only 109 staged samples from one hour of recording, with no confidence intervals and no per-class breakdown. The 'unseen' category pools 66 improper-zone, 7 unlawful-stop, and 18 improper-turn samples into a single binary decision, and the reported 79.10% accuracy and 0.88 F1 are suggestive but not statistically robust evidence that the model generalizes to unseen classes. Please report per-class performance, confidence intervals, and ideally frame-level temporal localization for the field-lab data.
  3. [Section IV.B, ShanghaiTech results] The paper states that TeG demonstrates 'better capability for identifying complex anomalies' and outperforms RTFM, but on ShanghaiTech TeG's 95.32% AUC is lower than RTFM's 97.21% and also lower than several other cited methods. The attribution of this gap to 'simpler anomalies' in ShanghaiTech is a post-hoc explanation without supporting analysis. The authors should either temper the superiority claim or provide an analysis, such as class-wise AUC, that explains the trade-off.
minor comments (5)
  1. [Section III, Eq. (1)] The notation in Eq. (1) is unclear: 'SL_i = P c_j' uses an undefined operator P, and the outer union over 32 segments appears to mix set notation with concatenation. Please rewrite the equation with explicit concatenation and indexing.
  2. [Section I and Section IV.D] There are several typos: 'real-word surveillance' should be 'real-world surveillance', 'future fusion' should be 'feature fusion', and Section IV.C contains 'and and'.
  3. [Section IV.A and III] The temporal granularity values (8, 32, 64), number of attention heads (4), and loss hyperparameters (m=100, k=3) are stated as empirically set, but no sensitivity analysis or ablation is provided. At minimum, please justify these choices or provide a short study.
  4. [References] The reference formatting is inconsistent: entries [12]–[14] and [15]–[16] use a different citation style from the rest of the bibliography. Please unify the formatting.
  5. [Table I] It is not stated explicitly whether the Table I results use the original UCF-Crime dataset or the extended 17-class version. Since Section IV.C states that the model is trained on the extended dataset, the protocol for Table I should be clarified to avoid ambiguity.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: TeG's accuracy claims rest on external benchmarks and an inherited external loss; the missing ablation weakens attribution but is not circular.

full rationale

The paper's derivation chain is: VST extracts features at three temporal granularities; MCA/MSA blocks fuse them; a 3-layer FCN classifier outputs segment scores; training uses RTFM's feature-magnitude and MIL ranking losses. Every evaluated claim (87.16% AUC on UCF-Crime, 95.32% AUC on ShanghaiTech, 84.57% AP on XD-Violence) is measured against public external benchmarks or a field-lab test set, not against the model's own fitted outputs. No equation in the paper defines a predicted value in terms of the fitted parameters in a way that would force the reported result. The temporal-granularity parameters (G=8, 32, 64 frames) are fixed empirical settings, and the attention fusion is a stated architectural choice, not a quantity fitted to the target metric. The only self-citation is Ref. [39] for the 'dangerous throwing' class added to the UCF-Crime extension; that citation supports a dataset class, not the core accuracy or the temporal-fusion claim. The absence of an ablation isolating the multi-granularity attention mechanism means the causal attribution of the gains to the proposed fusion is experimentally under-supported, but that is a completeness/attribution concern, not circularity per the criteria. No self-citation is load-bearing, no uniqueness theorem is imported from the authors, and no ansatz is smuggled in via citation: the RTFM loss [18] is external, the VST backbone [34] is external, and the attention mechanisms [25,35] are standard external components. Therefore the paper is self-contained against external benchmarks and receives a score of 0.

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

The central result rests on a pretrained VST backbone, the RTFM-style MIL loss, and empirically chosen granularities; no new entities are introduced. The main unstated load is that the chosen temporal scales and attention fusion, not the backbone or loss, cause the reported performance, which no ablation verifies.

free parameters (4)
  • Temporal granularity values (G) = 8, 32, 64 frames
    Set empirically in Section III to define short, medium, and long features; central to the method and not justified by ablation.
  • Attention head count = 4
    Empirically set in Section IV-A; not justified or ablated.
  • Loss margin m and top-k k = m=100, k=3
    Fixed in Section IV-A; inherited from the RTFM loss but still chosen constants.
  • Loss weighting factors lambda_FM, lambda_1, lambda_2 = not reported
    Weights in Eq. (4) are not given in Section IV-A, so the training loss is incompletely specified.
assumptions (3)
  • domain assumption Video-level labels are sufficient to supervise segment-level anomaly scores.
    The MIL/weak supervision framework from Sultani et al. [17] and RTFM [18] is adopted in Section III; no frame-level supervision is used.
  • domain assumption Video Swin Transformer features pretrained on Kinetics-400 transfer to surveillance anomaly detection.
    Section III uses VST [34] as a frozen feature extractor without reporting fine-tuning; this transfer is assumed.
  • domain assumption Anomalous segments have larger feature magnitudes than normal segments.
    The feature magnitude loss in Eq. (2) and top-k selection in Eq. (3) assume this separation; this is the training signal borrowed from [18].

how reviews work

0 comments
Cite this review

Pith. "Pith review of TeG: Temporal-Granularity Method for Anomaly Detection with Attention in Smart City Surveillance." pith.science (2026). https://pith.science/paper/7L5C2X3F

@misc{pith2026241111003,
  author       = {Pith},
  title        = {Pith review of: TeG: Temporal-Granularity Method for Anomaly Detection with Attention in Smart City Surveillance},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7L5C2X3F}},
  note         = {Machine review of arXiv:2411.11003}
}
read the original abstract

Anomaly detection in video surveillance has recently gained interest from the research community. Temporal duration of anomalies vary within video streams, leading to complications in learning the temporal dynamics of specific events. This paper presents a temporal-granularity method for an anomaly detection model (TeG) in real-world surveillance, combining spatio-temporal features at different time-scales. The TeG model employs multi-head cross-attention blocks and multi-head self-attention blocks for this purpose. Additionally, we extend the UCF-Crime dataset with new anomaly types relevant to Smart City research project. The TeG model is deployed and validated in a city surveillance system, achieving successful real-time results in industrial settings.

Figures

Figures reproduced from arXiv: 2411.11003 by the authors.

Figure 1
Figure 1. Workflow of the TeG model. The input video is split into 32 video segments. The Video Swin Transformer (VST) extracts features [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Anomaly visualizations and the detection scores/frame of the TeG model. The three sub-figures at the left display anomalies: [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 31 canonical work pages

  1. [1]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Georgescu, M.I., Barbalau, A., Ionescu, R.T., Khan, F.S., Popescu, M., Shah, M.: Anomaly detection in video via self-supervised and multi- task learning. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 12742–12752 (2021)

  2. [2]

    In: Proceedings of the IEEE/CVF inter- national conference on computer vision

    Liu, Z., Nie, Y ., Long, C., Zhang, Q., Li, G.: A hybrid video anomaly detection framework via memory-augmented flow reconstruction and flow-guided frame prediction. In: Proceedings of the IEEE/CVF inter- national conference on computer vision. pp. 13588–13597 (2021)

  3. [3]

    IEEE Access 9, 123977–123986 (2021)

    Yuan, H., Cai, Z., Zhou, H., Wang, Y ., Chen, X.: Transanomaly: Video anomaly detection using video vision transformer. IEEE Access 9, 123977–123986 (2021)

  4. [4]

    In: Proceedings of the 29th ACM International Conference on Multimedia

    Feng, X., Song, D., Chen, Y ., Chen, Z., Ni, J., Chen, H.: Convolutional transformer based dual discriminator generative adversarial networks for video anomaly detection. In: Proceedings of the 29th ACM International Conference on Multimedia. pp. 5546–5554 (2021)

  5. [5]

    In: 2022 26th International Conference on Pattern Recognition (ICPR)

    Lee, J., Nam, W.J., Lee, S.W.: Multi-contextual predictions with vision transformer for video anomaly detection. In: 2022 26th International Conference on Pattern Recognition (ICPR). pp. 1012–1018. IEEE (2022)

  6. [6]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Zaheer, M.Z., Mahmood, A., Khan, M.H., Segu, M., Yu, F., Lee, S.I.: Generative cooperative learning for unsupervised video anomaly detection. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 14744–14754 (2022)

  7. [7]

    In: Computer Vision–ACCV 2018: 14th Asian Conference on Computer Vision, Perth, Australia, December 2–6, 2018, Revised Selected Papers, Part III 14

    Akcay, S., Atapour-Abarghouei, A., Breckon, T.P.: Ganomaly: Semi- supervised anomaly detection via adversarial training. In: Computer Vision–ACCV 2018: 14th Asian Conference on Computer Vision, Perth, Australia, December 2–6, 2018, Revised Selected Papers, Part III 14. pp. 622–637. Springer (2019)

  8. [8]

    IEEE transactions on neural networks and learning systems 33(8), 3572–3586 (2021)

    Yu, J., Lee, Y ., Yow, K.C., Jeon, M., Pedrycz, W.: Abnormal event detec- tion and localization via adversarial event prediction. IEEE transactions on neural networks and learning systems 33(8), 3572–3586 (2021)

Show all 40 references
  1. [9]

    In: IEEE 10th International Conference on Signal Processing Proceedings

    Wang, S., Miao, Z.: Anomaly detection in crowd scene. In: IEEE 10th International Conference on Signal Processing Proceedings. pp. 1220–

  2. [10]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Wang, J., Cherian, A.: Gods: Generalized one-class discriminative subspaces for anomaly detection. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 8201–8211 (2019)

  3. [11]

    arXiv preprint arXiv:2105.14737 (2021)

    Kim, J.H., Kim, D.H., Yi, S., Lee, T.: Semi-orthogonal embed- ding for efficient unsupervised anomaly segmentation. arXiv preprint arXiv:2105.14737 (2021)

  4. [12]

    A Real-Time Deep Learning Approach for Real-World Video Anomaly Detection,

    S. Petrocchi, G. Giorgi, and M. G. Cimino, “A Real-Time Deep Learning Approach for Real-World Video Anomaly Detection,” in ARES 2021: The 16th International Conference on Availability, Reliability and Security. Association for Computing Machinery, 8 2021, pp. 1–9

  5. [13]

    Deep anomaly detection through visual attention in surveillance videos,

    N. Nasaruddin, K. Muchtar, A. Afdhal, and A. P. J. Dwiyantoro, “Deep anomaly detection through visual attention in surveillance videos,” Journal of Big Data , vol. 7, no. 1, pp. 1–17, 10 2020. [Online]

  6. [14]

    CNN features with bi-directional LSTM for real-time anomaly detection in surveillance networks,

    W. Ullah, A. Ullah, I. U. Haq, K. Muhammad, M. Sajjad, and S. W. Baik, “CNN features with bi-directional LSTM for real-time anomaly detection in surveillance networks,” Multimedia Tools and Applications, vol. 80, no. 11, pp. 16 979–16 995, 8 2020. [Online]

  7. [15]

    AnomalyNet: An Anomaly Detection Network for Video Surveillance,

    J. T. Zhou, J. Du, H. Zhu, X. Peng, Y . Liu, and R. S. M. Goh, “AnomalyNet: An Anomaly Detection Network for Video Surveillance,” IEEE Transactions on Information Forensics and Security , vol. 14, no. 10, pp. 2537–2550, 10 2019

  8. [16]

    Exploring background-bias for anomaly detection in surveillance videos,

    K. Liu and H. Ma, “Exploring background-bias for anomaly detection in surveillance videos,” in MM 2019 - Proceedings of the 27th ACM International Conference on Multimedia . Association for Computing Machinery, Inc, 10 2019, pp. 1490–1499. [Online]

  9. [17]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Sultani, W., Chen, C., Shah, M.: Real-world anomaly detection in surveillance videos. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 6479–6488 (2018)

  10. [18]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Tian, Y ., Pang, G., Chen, Y ., Singh, R., Verjans, J.W., Carneiro, G.: Weakly-supervised video anomaly detection with robust temporal feature magnitude learning. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 4975–4986 (2021)

  11. [19]

    arXiv preprint arXiv:2108.03825 (2021)

    Wu, J., Zhang, W., Li, G., Wu, W., Tan, X., Li, Y ., Ding, E., Lin, L.: Weakly-supervised spatio-temporal anomaly detection in surveillance video. arXiv preprint arXiv:2108.03825 (2021)

  12. [20]

    IEEE Transactions on Image Processing 30, 3513–3527 (2021)

    Wu, P., Liu, J.: Learning causal temporal relation and feature discrim- ination for anomaly detection. IEEE Transactions on Image Processing 30, 3513–3527 (2021)

  13. [21]

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

    Cho, M., Kim, M., Hwang, S., Park, C., Lee, K., Lee, S.: Look around for anomalies: Weakly-supervised anomaly detection via context-motion relational learning. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 12137–12146 (2023)

  14. [22]

    In: International Conference on Neural Information Processing

    Deshpande, K., Punn, N.S., Sonbhadra, S.K., Agarwal, S.: Anomaly detection in surveillance videos using transformer based attention model. In: International Conference on Neural Information Processing. pp. 199–

  15. [23]

    IEEE transactions on image processing 30, 4505–4515 (2021)

    Lv, H., Zhou, C., Cui, Z., Xu, C., Li, Y ., Yang, J.: Localizing anomalies from weakly-labeled videos. IEEE transactions on image processing 30, 4505–4515 (2021)

  16. [24]

    In: Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XXX

    Wu, P., Liu, J., Shi, Y ., Sun, Y ., Shao, F., Wu, Z., Yang, Z.: Not only look, but also listen: Learning multimodal violence detection under weak supervision. In: Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XXX

  17. [25]

    pp. 322–339. Springer (2020)

  18. [26]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Yan, S., Xiong, X., Arnab, A., Lu, Z., Zhang, M., Sun, C., Schmid, C.: Multiview transformers for video recognition. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 3333–3343 (2022)

  19. [27]

    In: European Conference on Computer Vision

    Li, G., Cai, G., Zeng, X., Zhao, R.: Scale-aware spatio-temporal relation learning for video anomaly detection. In: European Conference on Computer Vision. pp. 333–350. Springer (2022)

  20. [28]

    IEEE transactions on intelligent transportation systems 23(11), 21687–21699 (2022)

    Zhang, C., Li, G., Xu, Q., Zhang, X., Su, L., Huang, Q.: Weakly supervised anomaly detection in videos considering the openness of events. IEEE transactions on intelligent transportation systems 23(11), 21687–21699 (2022)

  21. [29]

    In: 2019 IEEE International Conference on Image Processing (ICIP)

    Zhang, J., Qing, L., Miao, J.: Temporal convolutional network with complementary inner bag loss for weakly supervised anomaly detection. In: 2019 IEEE International Conference on Image Processing (ICIP). pp. 4030–4034. IEEE (2019)

  22. [30]

    IEEE Signal Processing Letters 29, 2497–2501 (2022)

    Cao, C., Zhang, X., Zhang, S., Wang, P., Zhang, Y .: Adaptive graph convolutional networks for weakly supervised anomaly detection in videos. IEEE Signal Processing Letters 29, 2497–2501 (2022)

  23. [31]

    Pattern Recognition 77, 329–353 (2018)

    Carbonneau, M.A., Cheplygina, V ., Granger, E., Gagnon, G.: Multiple instance learning: A survey of problem characteristics and applications. Pattern Recognition 77, 329–353 (2018)

  24. [32]

    In: Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition

    Sapkota, H., Yu, Q.: Bayesian nonparametric submodular video partition for robust anomaly detection. In: Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition. pp. 3212–3221 (2022)

  25. [33]

    In: European Con- ference on Computer Vision

    Wu, J.C., Hsieh, H.Y ., Chen, D.J., Fuh, C.S., Liu, T.L.: Self-supervised sparse representation for video anomaly detection. In: European Con- ference on Computer Vision. pp. 729–745. Springer (2022)

  26. [34]

    arXiv preprint arXiv:1907.10211 (2019)

    Zhu, Y ., Newsam, S.: Motion-aware feature for improved video anomaly detection. arXiv preprint arXiv:1907.10211 (2019)

  27. [35]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Liu, Z., Ning, J., Cao, Y ., Wei, Y ., Zhang, Z., Lin, S., Hu, H.: Video swin transformer. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 3202–3211 (2022)

  28. [36]

    Advances in neural information processing systems 30 (2017)

    Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, Ł., Polosukhin, I.: Attention is all you need. Advances in neural information processing systems 30 (2017)

  29. [37]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Chen, Y ., Liu, Z., Zhang, B., Fok, W., Qi, X., Wu, Y .C.: Mgfn: Magnitude-contrastive glance-and-focus network for weakly-supervised video anomaly detection. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 37, pp. 387–395 (2023)

  30. [38]

    arXiv preprint arXiv:2311.15367 (2023)

    Zhou, Y ., Qu, Y ., Xu, X., Shen, F., Song, J., Shen, H.: Batchnorm- based weakly supervised video anomaly detection. arXiv preprint arXiv:2311.15367 (2023)

  31. [39]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Li, S., Liu, F., Jiao, L.: Self-training multi-sequence learning with trans- former for weakly supervised video anomaly detection. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 36, pp. 1395– 1403 (2022)

  32. [40]

    Electronic Imaging 35(9), 286–1–286–1 (2023)

    Kersten, I.P., Akdag, E., Bondarev, E., de With, P.H.: Detection of object throwing behavior in surveillance videos. Electronic Imaging 35(9), 286–1–286–1 (2023)

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.