The recognition of fighting behavior has been widely applied in various video-based domains. In the field of public safety, accurate and timely identification of fighting behavior in videos is crucial for making prompt decisions regarding such incidents. Among existing behavior recognition models, the SlowFast model has emerged as one of the most popular algorithms due to its dual-stream structure. However, its focus on local features limits its ability to extract global features, resulting in suboptimal classification accuracy. To address this issue, this paper proposes a fighting behavior recognition model incorporating an attention mechanism, which enhances the model's capability to identify behaviors more effectively. Compared to mainstream behavior recognition models, the proposed model demonstrates improved accuracy, offering valuable insights for addressing sudden incidents.

Youwei Jia. Detection of Fighting Behavior Based on an Improved SlowFast Architecture. Academic Journal of Computing & Information Science (2025), Vol. 8, Issue 2: 1-6. https://doi.org/10.25236/AJCIS.2025.080201.

[1] Feichtenhofer C, Fan H, Malik J, et al. Slowfast networks for video recognition[C]//Proceedings of the IEEE/CVF international conference on computer vision. 2019: 6202-6211.

[2] Zou M, Zhou Y, Jiang X, et al. Spatio-Temporal Behavior Detection in Field Manual Labor Based on Improved SlowFast Architecture[J]. Applied Sciences, 2024, 14(7): 2976.

[3] Vrskova R, Kamencay P, Hudec R, et al. A new deep-learning method for human activity recognition [J]. Sensors, 2023, 23(5): 2816. 

[4] Wang Z, Zheng J, Yang M, et al. Research on human behavior recognition in factory environment based on 3-2DCNN-BIGRU fusion network [J]. Signal, Image and Video Processing, 2025, 19(1): 1-12.