Facial emotion recognition using deep convolutional neural networks

Abstract

Human Emotions are the spontaneous reactions that occur to a human when they react to various situations. These emotions do not need conscious effort and they induce physiological changes in the muscles of the face while reflect in the facial expressions. The various Facial emotions that occur in the human’s face could include happy, sad, anger, disgust, fear, surprise and many more. These emotions that occur due to the inside feelings of a person play a key role in the non-verbal communication. There has been a lot of research that has been performed in the computer modelling of a human’s emotion, but the researches fall way behind the human vision. There have been various approaches to predict the human’s emotion but here we provide an approach where the human emotions could be predicted using deep convolutional neural networks. The FER-2013 database has been used for training in this project and the assessment of the proposed project provides good result and obtained encouraging accuracy for future projects.