Abstract
<div class="section abstract"><div class="htmlview paragraph">Driver state monitoring is a crucial technology for enhancing road safety and preventing human error-caused accidents in the era of autonomous vehicles. This paper presents CogniSafe, a comprehensive driver monitoring system that uses deep learning and computer vision methods to detect various types of driver distractions and fatigue. CogniSafe consists of four modules: <b><i>Driver anomaly detection and classification</i></b>: A novel two-phase network that proposes and recognizes driver anomalies, such as texting, drinking, and adjusting radios, using multimodal and multiview input. <b><i>Gaze estimation</i></b>: A video-based neural network that jointly learns head pose and gaze dynamics, achieving robust and efficient gaze estimation across different head poses. <b><i>Eye state analysis</i></b>: A multi-tasking CNN that encodes features from both eye and mouth regions, predicting the percentage of eye closure (PERCLOS) and the frequency of mouth opening (FOM). <b><i>Head pose estimation</i></b>: A CNN-based method that estimates the head pose from a single face image, providing additional information for driver attention assessment. CogniSafe integrates the outputs of each module into a specific driver status measurement and produces the driver's level of alertness on a scale from 0 to 5. The paper evaluates the performance of each module on benchmark datasets and discusses the practicality, technicality, and pivotal role of CogniSafe in the development and deployment of autonomous vehicles. The paper contributes to the field of driver monitoring by providing a novel and comprehensive system that covers the most common types of driver distractions and fatigue using deep learning and computer vision methods. The paper also provides a comprehensive overview of the current state-of-the-art and the future challenges of driver monitoring systems for a safer, more connected future of mobility.</div></div>