The monitoring of the driver’s state of concentration and the risk assessment of driving scenarios have received the majority of research interest in the area of automotive applications. In that situation, physiological tracking of the driver of the vehicle has shown to be a great non-invasive method of providing reliable and intelligent vehicle driving assistance. The authors of this work propose an advanced driving assistance system based on the usage of ad-hoc designed bio-inspired sensing platform which samples the driver’s PhotoPlethysmoGraphic (PPG) signal, properly correlated with the respective level of subject attention. The perceptual analysis of the driver’s attention level is certainly a robust and non-invasive approach for drowsiness monitoring. An innovative downstream deep architecture will be used to adequately process the so-collected driver’s PPG signal by reconstructing the corresponding attention level. Extensive scientific literature confirmed that the PPG signal can be adequately used to reconstruct the activation of the autonomic nervous system of the monitored subject and therefore of the correlated level of attention managed by the activation of the sympathetic or parasympathetic nerve. Additionally, a video saliency analysis-based external intelligent automotive-grade vision-based system will be in charge of determining the amount of danger associated with the driving situation. The gathered experiment results attested to the usefulness of the whole pipeline proposal.
Advanced Intelligent deep learning-based system for Robust Driving Assistance
2023-07-17
2254798 byte
Conference paper
Electronic Resource
English
Advanced driving assistance system and advanced driving assistance method
European Patent Office | 2021
|