Advanced Temporal Dilated Convolutional Neural Network for a Robust Car Driver Identification

Abstract

The latest generation cars are often equipped with advanced driver assistance systems usually known as ADAS (Advanced Driver Assistance Systems). These systems are able to assist the car driver leveraging several levels of automation. It is therefore essential to adapt the ADAS technology to the car driver's identity in order to personalize the provided assistance services. For these reasons, such car driver profiling algorithms have been developed by scientific community. The algorithm herein proposed is able to recognize the driver's identity with an accuracy close to 99% thanks to ad'hoc specific analysis of the driver's PhotoPlethysmoGraphic (PPG) signal. In order to rightly identify the driver profile, the proposed approach uses a 1D Dilated Temporal Convolutional Neural Network architecture to learn the features of the collected driver's PPG signal. The proposed deep architecture is able to correlate the specific PPG features with subject identity enabling the car ADAS services associated to the recognized identity. Extensive validation and testing of the developed pipeline confirmed its reliability and effectiveness.

Deep Neuro-Vision Embedded Architecture for Safety Assessment in Perceptive Advanced Driver Assistance Systems: The Pedestrian Tracking System Use-Case

“…Another automotive issue we have addressed is the robust identification of the car driver. Through adhoc intelligent pipeline based on the usage of PPG signal and deep network, we have designed the so called "physiological fingerprint" of the driver used for an effective identification (Rundo et al, 2020d). Therefore the proposed method can be generalized and applied in various automotive scenarios in which it is necessary to characterize the level of attention of the driver or in all the use-cases that deal with driving safety.…”