In this proposed system Convolution Neural Network (CNN) is used to detect human facial emotion and for the better performance outcome.
Keras is used as a deep learning API which helps to design deep learning model. Standard Data FER2013 is used as the dataset which consists of 30,219 images which is further used for the purpose of training (24,282) and validation (5937) of images. This project helps to detect a face within a boundary of 48*48 pixel and detects the emotion for the inputted images. The basic five emotions of human beings are Happy, Sad, Angry, Disgust, Fearful, Neutral and Surprise.
