Deep Neural Network for Facial Emotion Recognition System

Facial emotion recognition is a challenging task to capture and analyze. This paper presents an intelligent approach to detect different facial expressions of a person. In this context, a noval deep learning vgg 19 convolutional neural network architecture is used. Seven different emotions (anger, disgust, fear, happiness, sadness, surprise, and neutral) are considered to recognize emotions. The FER 2013 dataset contains 35,887 different images with these emotions. Results were obtained by using various activation functions. A comparative study of various activation functions has been done.