Emotion detection and face recognition of drivers in autonomous vehicles in IoT platform

Abstract

The Internet of Things (IoT) has been used in many fields, such as medical treatment, video monitoring, and transport. In these regions, the rapid adoption and development of IoT produce a large amount of evidence. For instance, IoT devices such as cameras capture the different facial emotions of drivers in Autonomous driving systems. In completely autonomous vehicles, drivers have trouble in takeover transfers as they become disconnected from the actual aspect of driving. Factors affecting takeover effectiveness, such as lead time and the participation of non-driving-related tasks, have been concentrated. Nevertheless, considering the vital role of emotions, human communication and manual driving affect drivers' takeover efficiency. Face identity is essential for emotional sensitivity detection for drivers in autonomous vehicles. Automated and intelligent face recognition (FR) devices are highly accurate in a comfortable condition and unregulated, with poor reliability in autonomous vehicles. Artificial Intelligence (AI) can significantly perceive and express feelings in well-being and similar fields. This study suggests an optimized IoT architecture that facilitates the physiological signal with wireless transmission to the database management center. Face Recognition and Emotion Detection based on IoT (FRED-IoT) has been proposed to track drivers' emotional and face recognition in autonomous vehicles. A low delay of 2 milliseconds is achieved in the proposed IoT Protocols. In contrast with cutting-edge technology, FRED-IoT improves reliability and achieves a high rating (F-score) of 96%.