Alertness Anticipation Innovations in Driver Safety Predictions

• Published in: 2023 IEEE International Conference on ICT in Business Industry & Government (ICTBIG) pp. 1 - 5
• Main Authors: S, Priyanka, S, Keerthika, J, Arunesh, S, Priyadharshini, S, Sanjay, C, Anudeep
• Format: Conference Proceeding
• Language: English
• Published: IEEE 08-12-2023
• Subjects: CNN; Deep Learning; KNN; Machine Learning; Machine learning algorithms; Random Forest; Real-time systems; RNN; Road safety; Sleep; Support vector machines; SVM; Technological innovation; Time measurement
• DOI: 10.1109/ICTBIG59752.2023.10456021

Drowsy driving is a major contributor to traffic accidents and a major risk to public safety. The development of driver sleepiness detection systems has found success in recent years with the use of machine learning techniques. This work proposes a comprehensive approach to machine learning algorithms-such as Support Vector Machines (SVM), Random Forest, Decision Tree, and K-Nearest Neighbours (KNN)-for the detection of driver sleepiness. In this paper, investigate how these algorithms perform when trained on a picture dataset. Using a camera mounted in the car, the suggested system first takes live pictures of the driver's face. After that, facial landmarks are taken out of these pictures and used as input data for machine learning models. To identify the patterns linked to drowsiness, each algorithm is trained using a labelled dataset that includes both alert and sleepy facial expressions. Using cross-validation approaches, the performance of each algorithm is assessed in terms of accuracy, precision, recall, and F1-score. By informing sleepy drivers in real time so they can take the appropriate safety measures or take a break, the suggested system has the potential to improve road safety. An important way to assess the suitability of various machine learning algorithms for driver sleepiness detection applications is to compare them. To improve the system's accuracy and resilience in real-world situations, more research can concentrate on incorporating other sensor data, such as eye movement patterns or steering habits.