Driver Emotion Recognition for Enhancement of Human Machine Interface in Vehicles

Emotion recognition of a driver is a crucial task in vehicles. Emotions have an adverse influence on driving behavior. Emotions like anger, fear, and sadness have a negative impact on driver response time and may lead to fatal accidents. This paper considers five specific kinds of emotions labelled anger, fear, happy, neutral, and sadness, that may occur in a driver. This method fuses both LBP and facial landmark features to detect emotions. The supervised machine learning algorithm, Support Vector Machine (SVM) is used for the classification of different emotions. Performance on extended Cohn-Kanade dataset is obtained, exhibited and analyzed. With this proposed method, we obtained an accuracy of 86.7%.