Authors:
C. N. Sangeetha, K. Vinoth Kumar, M. Arunadevi Thirumalraj, S. Venkatasubramanian, Jouma Ali Al-Mohamad

Addresses:
Department of Electrical and Electronics Engineering, New Horizon College of Engineering, Bengaluru, Karnataka, India. Department of Computer Science and Engineering, Karunya Institute of Technology and Science, Coimbatore, Tamil Nadu, India. Department of Computer Science and Business Management, Saranathan College of Engineering, Trichy, Tamil Nadu, India. Department of Computer Science and Business Systems, Saranathan College of Engineering, Trichy, Tamil Nadu, India. Department of Computer and Mobile Communications Engineering, Al-Shahbaa Private University, Aleppo, Aleppo Governorate, Syria.

Abstract:

In 20% of fatal car accidents, drivers are tired. One of the most promising and commonly available technologies is vehicle driver fatigue monitoring. This item will determine if video-based analysis and machine learning can detect tiredness. Method development and evaluation were based on the experiences of 13 Swedish freeway drivers. Each driver participated in two 90-minute sessions, one during daylight (low sleepiness) and one at midnight (high sleepiness). The KSS sleepiness detection model outputs an alert or weary (KSS 7-9) or a score of 1-9. For driver tiredness prediction, we recommend BH-XGBoost, which uses Bayesian hyperparameter optimisation. Feature selection improves the model's classification accuracy. HBC enhances the BRO algorithm. For participant connection, the BRO method uses the chicken swarm optimisation (CSO) algorithm. An elite player update method that combines random exploration and directed updating within a confined zone may enhance growth. Real-time drowsiness detection can help control fatigue and prevent driving accidents by identifying severe indicators of weariness. With 100 iterations, the suggested model achieved 81% accuracy and 91% precision, and 79% accuracy and 83% precision without a feature selection model.

Keywords: Karolinska Sleepiness Scale; Driver Fatigue; Chicken Swarm Optimization; Battle Royale Optimization; Bayesian Hyperparameter Optimization; Elite Random Guidance; Gaussian Time Domain Network.

Received on: 13/11/2024, Revised on: 07/02/2025, Accepted on: 19/03/2025, Published on: 07/09/2025

DOI: 10.69888/FTSCS.2025.000476

FMDB Transactions on Sustainable Computing Systems, 2025 Vol. 3 No. 3, Pages: 176-191

• Views : 71
• Downloads : 9

Download PDF