Chinese Journal of Liquid Crystals and Displays, Volume. 40, Issue 4, 617(2025)

Fatigue driving detection based on improved YOLOv8n-Pose

Zhongqi CAI1,2, Shanling LIN1,2, Jianpu LIN1,2, Shanhong LÜ1,2, Zhixian LIN1,2、*, and Tailiang GUO2
Author Affiliations
  • 1School of Advanced Manufacturing,Fuzhou University,Quanzhou 362251,China
  • 2Fujian Science and Technology Innovation Laboratory for Photoelectric Information,Fuzhou 350116,China
  • show less
    Figures & Tables(15)
    YOLOv8n-Pose network structure
    Diagram of Ghost convolution operation
    C3Ghost module structure
    Diagram of GS convolution operation
    VoV-GSCSP module structure
    GNSC detection head structure
    SEAM attention module
    Improved YOLOv8n-Pose network structure
    Face key points
    Comparison images before and after improvement.(a)Label;(b)YOLOv8n-Pose;(c)Ours.
    Fatigue driving detection.(a)Normal;(b)Closing eyes for too long;(c)Closing eyes for too long and yawning frequently;(d)Yawning frequently.
    • Table 1. Experimental environment configuration

      View table
      View in Article

      Table 1. Experimental environment configuration

      名称环境配置
      操作系统Windows 11 64位
      GPUNVIDIA GeForce RTX 2080Ti
      内存16 GB
      PythonPython3.8.17版本
      深度学习框架Pytorch2.0.0
    • Table 2. Ablation experiment

      View table
      View in Article

      Table 2. Ablation experiment

      GhostSlim-neckGNSCSEAMmAP@0.5Parm/MGFLOPsFPS
      ××××0.9563.29.2635
      ×××0.9582.57.2668
      ××0.9622.36.4673
      ×0.9601.54.4742
      0.9651.64.5683
    • Table 3. Comparison experiments

      View table
      View in Article

      Table 3. Comparison experiments

      ModelmAP@0.5Parm/MGFLOPsFPS
      HigherHRNet0.94563.8154.3-
      DEKR0.95365.7141.5-
      YOLOv5n-Pose0.9514.58.4629
      YOLOv7n-Pose0.9595.410.2525
      YOLOv8n-Pose0.9563.29.2635
      本文算法0.9651.64.4683
    • Table 4. Fatigue driving detection

      View table
      View in Article

      Table 4. Fatigue driving detection

      视频编号实际闭眼次数检测闭眼次数实际哈欠次数检测哈欠次数实际疲劳次数算法判断疲劳次数疲劳判定准确率/%
      143413344100
      2125119559888.8
      35551447685
      431293344100
    Tools

    Get Citation

    Copy Citation Text

    Zhongqi CAI, Shanling LIN, Jianpu LIN, Shanhong LÜ, Zhixian LIN, Tailiang GUO. Fatigue driving detection based on improved YOLOv8n-Pose[J]. Chinese Journal of Liquid Crystals and Displays, 2025, 40(4): 617

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category:

    Received: Jul. 12, 2024

    Accepted: --

    Published Online: May. 21, 2025

    The Author Email: Zhixian LIN (lzx2005000@163.com)

    DOI:10.37188/CJLCD.2024-0192

    Topics