DARTS Meets Ants: A Hybrid Search Strategy for Optimizing KAN-Based 3D CNNs for Violence Recognition in Video
Source abstract
Abstract
The optimization capabilities of Kolmogorov–Arnold Networks (KANs) remain largely unexplored, which has limited their practical use in video anomaly recognition compared to conventional 3D-CNNs. To address this gap, we introduce a novel hybrid optimization framework that combines a Minimax Ant System (MMAS) for hyperparameter selection with a modified DARTS strategy for adaptive tuning of the 3D KAN architecture. Unlike existing approaches, our method simultaneously optimizes both learning dynamics and architectural configurations, enabling KANs to better exploit their expressive power in spatiotemporal feature extraction. Applied to a three-class video dataset, the proposed approach improved model accuracy to 87%, surpassing the performance of a standard 3D-CNN by 6%.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Anomaly Detection Techniques and Applications · Human Pose and Action Recognition · Video Surveillance and Tracking Methods
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Institute of Information and Computational Technologies, Almaty 050010, Kazakhstan
Department of Computer Science, Al-Farabi Kazakh National University, Almaty 050040, Kazakhstan
Funding
The Committee of Science of the Ministry of Science and Higher Education of the Republic of Kazakhstan
References (29)
Show source-provided references
1. Park, J.H., Mahmoud, M., and Kang, H.S. (2024). Conv3D-based video violence detection network using optical flow and RGB data. Sensors, 24.10.3390/s24020317 ↗
2. Karim, A., Razin, J.I., Ahmed, N.U., and Alam, T. (2021). An automatic violence detection technique using 3D convolutional neural network. Sustainable Communication Networks and Application: Proceedings of ICSCN 2020, Springer.
3. Analysis of anomalous crowd behavior by employing pre-trained efficient-X3D net for violence detection10.1007/s12046-025-02686-1 ↗
4. Anomaly recognition from surveillance videos using 3D convolution neural network10.1007/s11042-021-10570-3 ↗
5. Liu, Z., Wang, Y., Vaidya, S., Ruehle, F., Halverson, J., Soljačić, M., and Tegmark, M. (2024). Kan: Kolmogorov-arnold networks. arXiv.
6. Long, Q., Wang, B., Xue, B., and Zhang, M. (2025). A Genetic Algorithm-Based Approach for Automated Optimization of Kolmogorov-Arnold Networks in Classification Tasks. arXiv.
7. Sen, A., Mazumder, A.R., Dutta, D., Sen, U., Syam, P., and Dhar, S. (2023, January 13–15). Comparative evaluation of metaheuristic algorithms for hyperparameter selection in short-term weather forecasting. Proceedings of the 15th International Joint Conference on Computational Intelligence—ECTA, Rome, Italy.10.5220/0012187300003595 ↗
8. Liu, H., Simonyan, K., and Yang, Y. (2018). DARTS: Differentiable architecture search. arXiv.
9. A multi-stream CNN for deep violence detection in video sequences using handcrafted features10.1007/s00371-021-02266-4 ↗
10. Crimenet: Neural structured learning using vision transformer for violence detection10.1016/j.neunet.2023.01.048 ↗
11. Shanmughapriya, M., Gunasundari, S., Fenitha, J.R., and Sanchana, R. (2022, January 14–16). Fight detection in surveillance video dataset versus real time surveillance video using 3DCNN and CNN-LSTM. Proceedings of the 2022 International Conference on Computer, Power and Communications (ICCPC), Chennai, India.
12. Comparison of three evolutionary algorithms: GA, PSO, and DE
13. Sen, A., Mazumder, A.R., and Sen, U. (2023, January 5–8). Differential evolution algorithm based hyper-parameters selection of transformer neural network model for load forecasting. Proceedings of the 2023 IEEE Symposium Series on Computational Intelligence (SSCI), Mexico City, Mexico.10.1109/SSCI52147.2023.10371846 ↗
14. Deep multi-model fusion for human activity recognition using evolutionary algorithms
15. Abdelmoaty, A.M., and Ibrahim, I.I. (2024). Comparative Analysis of Four Prominent Ant Colony Optimization Variants: Ant System, Rank-Based Ant System, Max-Min Ant System, and Ant Colony System. arXiv.
16. A comprehensive survey of neural architecture search: Challenges and solutions
17. Latypov, V., and Hvatov, A. (2025, January 24–28). Exploring convolutional KAN architectures with NAS. Proceedings of the First Conference of Mathematics of AI, Sochi, Russia.
18. Safety helmet monitoring on construction sites using YOLOv10 and advanced transformer architectures with surveillance and body-worn cameras10.1061/JCEMD4.COENG-16760 ↗
19. Tang, H., Liu, J., Yan, S., Yan, R., Li, Z., and Tang, J. (November, January 29). M3net: Multi-view encoding, matching, and fusion for few-shot fine-grained action recognition. Proceedings of the 31st ACM International Conference on Multimedia, Ottawa, ON, Canada.
20. Vaca-Rubio, C.J., Blanco, L., Pereira, R., and Caus, M. (2024, January 8–12). Kolmogorov-arnold networks (KANs) for time series analysis. Proceedings of the 2024 IEEE Globecom Workshops (GC Wkshps), Cape Town, South Africa.10.1109/GCWkshp64532.2024.11100692 ↗
21. Cheon, M. (2024). Demonstrating the efficacy of kolmogorov-arnold networks in vision tasks. arXiv.
22. Genet, R., and Inzirillo, H. (2024). Tkan: Temporal kolmogorov-arnold networks. arXiv.10.2139/ssrn.4825654 ↗
23. Drokin, I. (2024). Kolmogorov-Arnold Convolutions: Design Principles and Empirical Studies. arXiv.
24. Szeghalmy, S., and Fazekas, A. (2023). A comparative study of the use of stratified cross-validation and distribution-balanced stratified cross-validation in imbalanced learning. Sensors, 23.10.3390/s23042333 ↗
25. Vrskova, R., Hudec, R., Kamencay, P., and Sykora, P. (2022). Human activity classification using the 3DCNN architecture. Appl. Sci., 12.10.3390/app12020931 ↗
26. Bodner, A.D., Tepsich, A.S., Spolski, J.N., and Pourteau, S. (2024). Convolutional kolmogorov-arnold networks. arXiv.
27. ML-MMAS: Self-learning ant colony optimization for multi-criteria journey planning10.1016/j.ins.2022.07.150 ↗
28. An architecture entropy regularizer for differentiable neural architecture search10.1016/j.neunet.2022.11.015 ↗
29. Ali, Y.A., Awwad, E.M., Al-Razgan, M., and Maarouf, A. (2023). Hyperparameter search for machine learning algorithms for optimizing the computational complexity. Processes, 11.10.3390/pr11020349 ↗