1. A lightweight fatigue driving detection method based on facial features.
- Author
-
Zhu, Jun-Wei, Ma, Yan-E, Xia, Jia, and Zhou, Xiao-Gang
- Abstract
Fatigue driving is a significant factor in traffic accidents, underscoring the importance of driver fatigue detection. Existing fatigue detection methods are susceptible to factors such as acquisition angle and face occlusion, and a large number of parameters make them unsuitable for deployment on embedded devices. To address these issues, this paper proposes a lightweight fatigue driving detection method based on facial features. Firstly, a lightweight face detection method is presented, which introduces a lightweight backbone network to extract effective features. Subsequently, a lightweight facial landmark detection method is proposed to enhance positioning accuracy by modifying the structure of the model and the loss function, obtaining 98 landmarks of the driver, and allowing for the calculation of fatigue characterization parameters for the eyes, mouth, and head. Finally, we combine the extracted fatigue characteristic parameters to comprehensively assess the driver's fatigue state. Experimental results demonstrate that our method exhibits low model complexity, fulfilling the real-time and accuracy requirements. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF