Road safety is crucial to prevent traffic deaths and injuries of drivers, passengers, and pedestrians. Various regulations and policies have been proposed to aim at reducing the number of traffic deaths and injuries. However, these figures have remained steady in recent decade. There has been an increasing number of research works on the prediction of driver status which gives warning before undesired status, for instance drowsiness and stress. In this paper, a long short-term memory networks is proposed for generic design of driver drowsiness prediction and driver stress prediction models using electrocardiogram (ECG) signals. The proposed model achieves sensitivity, specificity, and accuracy of 71.0–81.1%, 72.9–81.9%, and 72.2–81.5%, respectively, for driver drowsiness prediction. They are 68.2–79.3%, 71.6–80.2%, and 70.8–79.7%, for driver stress prediction. The results have demonstrated the feasibility of generic model for both drowsiness and stress prediction. Future research directions have been shared to enhance the model performance.
Long Short-Term Memory Networks for Driver Drowsiness and Stress Prediction
Advs in Intelligent Syst., Computing
International Conference on Intelligent Computing & Optimization ; 2020 ; Koh Samui, Thailand December 17, 2020 - December 18, 2020
08.02.2021
11 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Long Short Term Memory for Driver Intent Prediction
British Library Conference Proceedings | 2017
|Long short term memory for driver intent prediction
IEEE | 2017
|IEEE | 2020
|Springer Verlag | 2017
|