A lightweight video violence detection method based on multi-scale temporal pyramid networks
DOI:
https://doi.org/10.66238/fsrma94Keywords:
Violence detection, Temporal pyramid network, Multi scale modeling, Temporal attentionAbstract
Video violence detection plays a critical role in public safety surveillance and online content moderation. This paper proposes a Multi-Scale Temporal Pyramid Network that employs three parallel temporal pyramid branches with strides of 1, 2, and 4 to explicitly model fine-grained, intermediate, and coarse temporal dynamics. The model adopts a lightweight MobileNetV3-Large backbone for spatial feature extraction, followed by spatial-to-temporal projection to generate compact token sequences. Three temporal depth-wise separable residual blocks independently process each temporal scale, while a temporal scale attention fusion module adaptively integrates multi-scale representations. Learnable temporal attention pooling then compresses variable-length sequences into a compact video-level descriptor for classification. Systematic experiments on three public benchmark datasets: RWF-2000, Hockey Fights, and Real Life Violence Situations, demonstrate that MPPN achieves competitive detection performance, validating the effectiveness of explicit multi-scale temporal modeling for practical video violence detection.
References
[1] Huszar, V. D., Adhikarla, V. K., Negyesi, I., & Krasznay, C. (2023). Toward fast and accurate violence detection for automated video surveillance applications. IEEE Access, 11, 18772-18793.
[2] Binxu, W. A. N. G., & ZHANG, X. (2023). An Efficient Violence Detection Method Based on Temporal Attention Mechanism. Instrumentation, 10(2).
[3] Ullah, F. U. M., Obaidat, M. S., Ullah, A., Muhammad, K., Hijji, M., & Baik, S. W. (2023). A comprehensive review on vision-based violence detection in surveillance videos. ACM Computing Surveys, 55(10), 1-44.
[4] Chen, Y., Li, Y., Li, S., Lv, S., & Lin, F. (2025). DualCascadeTSF-MobileNetV2: a lightweight violence behavior recognition model. Applied Sciences, 15(7), 3862.
[5] Liu, P., Yan, S., Li, C., & Nyamasvisva, T. E. (2025, September). Lightweight 3D Convolutional Neural Network with SE Attention for Campus Violence Recognition. In 2025 7th International Conference on Internet of Things, Automation and Artificial Intelligence (IoTAAI) (pp. 240-243). IEEE.
[6] Vosta, S., & Yow, K. C. (2023). KianNet: A violence detection model using an attention-based CNN-LSTM structure. IEEE Access, 12, 2198-2209.
[7] Rendón-Segador, F. J., Álvarez-García, J. A., Salazar-González, J. L., & Tommasi, T. (2023). Crimenet: Neural structured learning using vision transformer for violence detection. Neural networks, 161, 318-329.
[8] Islam, Z., Rukonuzzaman, M., Ahmed, R., Kabir, M. H., & Farazi, M. (2021, July). Efficient two-stream network for violence detection using separable convolutional lstm. In 2021 International Joint Conference on Neural Networ
[9] Bai, Y., Yang, S., Zhang, Z., Jiang, P., Tang, L., & Yu, X. (2024, December). Recognition of Campus Violence Based on Transformer Model Using Thermal Infrared Videos. In 2024 5th International Conference on Information Science and Education (ICISE-IE) (pp. 238-242). IEEE.
[10] Bermejo Nievas, E., Deniz Suarez, O., Bueno García, G., & Sukthankar, R. (2011). Violence detection in video using computer vision techniques. In Computer Analysis of Images and Patterns: 14th International Conference, CAIP 2011, Seville, Spain, August 29-31, 2011, Proceedings, Part II 14 (pp. 332-339). Springer Berlin
[11] Cheng, M., Cai, K., & Li, M. (2019). RWF-2000: An Open Large Scale Video Database for Violence Detection. http://arxiv.org/abs/1911.05913
[12] Soliman, M. M., Kamal, M. H., Nashed, M. A. E. M., Mostafa, Y. M., Chawky, B. S., & Khattab, D. (2019, December). Violence recognition from videos using deep learning techniques. In 2019 ninth international conference on intelligent computing and information systems (ICICIS) (pp. 80-85). IEEE.
[13] Khan, H., Yuan, X., Qingge, L., & Roy, K. (2025). Violence Detection From Industrial Surveillance Videos Using Deep Learning. IEEE Access, 13, 15363–15375. https://doi.org/10.1109/ACCESS.2025.3531213
[14] Dündar, N., Keçeli, A. S., Kaya, A., & Sever, H. (2024). A shallow 3D convolutional neural network for violence detection in videos. Egyptian Informatics Journal, 26, 100455.
[15] Xia, X., Wu, H., & Yang, C. (2021). Violence detection with two-stream neural network based on C3D. International Journal of Cognitive Informatics and Natural Intelligence (IJCINI), 15(4), 1-17.
Toor, M. J., Shabbir, S., Rauf, A., Nisar, A., Safdar, S., & Zaib-Un-Nisa. (n.d.). The role of structured-communication-technique in developing pragmatic language capability among ASD-children. https://doi.org/10.63878/cjssr.v3i2.785
Toor, M. J., Mehmood, H., Yasmeen, A., Honey, Q., Waseem, M., & Wasif, M. (n.d.). Exploring the effect of verbal encouraging phrases (VEP) on Chinese language learners’ academic achievement at university level.
Toor, M. J., Arafat, M., Ayoub, M., Ali, M., & Alessi Longa, F. E. (n.d.). From coherent self to fragmented persona: Digital identity fragmentation and academic self-concept among Generation Z, with comparative insights from Generation Alpha. https://doi.org/10.63878/cjssr.v4i2.2318
Katsitadze, N., Kharadze, N., & Pirtskhalaishvili, D. (2025). The impact of social media on students’ effective time management. Baltic Journal of Economic Studies, 11(1), 200–210. https://doi.org/10.30525/2256-0742/2025-11-1-200-210
Toor, M. J., Zafar, S., Omer, F., Sarwar, M., & Hussain, A. K. (n.d.). Teachers’ perspectives on effectiveness of translanguaging in enhancing English language learning at secondary school level.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Liu Pan, Liu Guangbin, Tadiwa Elisha Nyamasvisva
This work is licensed under a Creative Commons Attribution 4.0 International License.
License Terms (Recommended – CC BY 4.0)
All articles published in Fundamental Scientific Reports in Multidisciplinary Areas are licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0).
This license permits unrestricted use, distribution, and reproduction in any medium, provided the original author(s) and source are properly cited. Users are allowed to copy, redistribute, remix, transform, and build upon the material for any purpose, including commercial use, as long as appropriate credit is given.
Authors retain copyright of their work while granting the journal the right of first publication. The journal provides immediate open access to its content on the principle that making research freely available supports a greater global exchange of knowledge.
For more information about this license, visit:
https://creativecommons.org/licenses/by/4.0/
