Driver drowsiness estimation using EEG signals with a dynamical encoder–decoder modeling framework

Abstract

Drowsiness is a leading cause of accidents on the road as it negatively affects the driver’s ability to safely operate a vehicle. Neural activity recorded by EEG electrodes is a widely used physiological correlate of driver drowsiness. This paper presents a novel dynamical modeling solution to estimate the instantaneous level of the driver drowsiness using EEG signals, where the PERcentage of eyelid CLOSure (PERCLOS) is employed as the ground truth of driver drowsiness. Applying our proposed modeling framework, we find neural features present in EEG data that encode PERCLOS. In the decoding phase, we use a Bayesian filtering solution to estimate the PERCLOS level over time. A data set that comprises 18 driving tests, conducted by 13 drivers, has been used to investigate the performance of the proposed framework. The modeling performance in estimation of PERCLOS provides robust and repeatable results in tests with manual and automated driving modes by an average RMSE of 0.117 (at a PERCLOS range of 0 to 1) and average High Probability Density percentage of 62.5%. We further hypothesized that there are biomarkers that encode the PERCLOS across different driving tests and participants. Using this solution, we identified possible biomarkers such as Theta and Delta powers. Results show that about 73% and 66% of the Theta and Delta powers which are selected as biomarkers are increasing as PERCLOS grows during the driving test. We argue that the proposed method is a robust and reliable solution to estimate drowsiness in real-time which opens the door in utilizing EEG-based measures in driver drowsiness detection systems.

A Novel Approach for Simulation of Automotive Radar Sensors Designed for Systematic Support of Vehicle Development

“…In addition, the authors in [ 35 ] introduce two DL-based frameworks for electroencephalography-based (EEG) human intention recognition applied to a brain–computer interface (BCI). In the automotive domain, using the dynamic encoder–decoder modelling framework, an EEG signal-based driver drowsiness estimation application was introduced in [ 36 ]. Coming back to simulating perception sensors, a radar data simulation using deep generative networks was presented by the authors in [ 37 ].…”

Association of Visual-Based Signals with Electroencephalography Patterns in Enhancing the Drowsiness Detection in Drivers with Obstructive Sleep Apnea

“…Research highlights EEG measurements as highly effective in promptly identifying drowsiness onset, surpassing both behavior-based and vehicle-based systems [ 24 , 25 , 26 ]. Although, behavior-based system lags subtly behind EEG measures in identifying drowsiness onset, detecting early signs such as eye-blinking linked to drowsiness before any lateral vehicle displacement occurs [ 27 ]. Since vehicle-based systems issue alerts later in the initial drowsiness phase, potentially limiting accident prevention opportunities, relying solely on this technique is not advisable.…”

“…PERCLOS, a behavior-based signal approved by NHTSA for drowsiness detection independently, measures the duration of eyes at least 80 percent closed within a minute [ 28 ]. It can be calculated using built-in algorithms in eye-tracking systems (ETSs) like SmartEye [ 27 ] or via image processing from recorded facial videos [ 18 , 19 ]. However, limitations in image processing include challenges with video/image quality, eyewear interference, varying lighting, and head movement, impacting performance [ 24 , 26 , 27 ].…”

“…It can be calculated using built-in algorithms in eye-tracking systems (ETSs) like SmartEye [ 27 ] or via image processing from recorded facial videos [ 18 , 19 ]. However, limitations in image processing include challenges with video/image quality, eyewear interference, varying lighting, and head movement, impacting performance [ 24 , 26 , 27 ]. Addressing these constraints is crucial, as relying solely on this technique may lack reliability.…”

“…Several studies have integrated PERCLOS with additional behavioral parameters, such as average eye closure speed [ 29 ], head movement [ 30 ], and yawning episodes [ 31 ], enhancing the effectiveness of drowsiness detection. Furthermore, study [ 27 ] correlated PERCLOS-based drowsiness with neural patterns, revealing increased theta and delta powers as PERCLOS escalated during driving tests. Study [ 28 ] utilized photoplethysmography imaging (PPGI) to derive heart rate variability (HRV) and LF/HF ratio, achieving 92.5% accuracy by correlating these HRV-derived parameters with PERCLOS measurements.…”