Although the driverβs emotion has been studied in the different driving environments (such as city and highway), understanding what eye metrics and facial expressions correspond to specific emotion and behavior based on subjective and biosensor data to study emotion in depth is not well researched in previous studies. Using an eye-integrated human-in-the-loop (HTIL) simulation experiment, we studied how driversβ facial expressions and ocular measurements relate to emotions. We found that the driving environment could significantly affect driversβ emotions, which is evident in their facial expressions and eye metrics data. In addition, such outcomes provide knowledge to human-computer-interaction (HCI) practitioners on designing emotion recognition systems in cars to have a robust understanding of the driversβ emotions and help progress multimodal emotion recognition.
