A Deep Learning Approach for Real Time Facial Emotion Recognition

Abstract

Facial emotion recognition is a trending technique for understanding the emotions. The current study focusses of studying the deep learning techniques for understanding human emotion through facial emotions. The proposed model includes applying Convolutional Neural Network based model to understand facial emotion recognition. Three facial datasets-LFW, The Extended Yale Face Database B, Google Facial expression comparison dataset were used to calculate the six discrete emotions namely as happy, sad, angry, surprise, bore, and disgust for emotion recognition. The results of the experiment are compared with existing facial emotion recognition techniques and results shows that proposed model has attained an accuracy of ~93% which is quite better than existing research.