Face Expression Recognition Classification Using Deep Learning

This study seeks to design an intelligent system for face expression classification through the application of deep learning methods. Face images are analyzed to extract salient features and pattern them into facial expressions like happiness, sadness, anger, surprise, disgust and fear with a neutral face expression. A Convolutional Neural Network (CNN) framework was used because it is very effective in the analysis of images and in recognizing patterns. The model is trained from typical facial expression databases and tested with performance indicators like accuracy and error rate. The outcomes of the study reveal that the proposed model is highly accurate in the classification of feelings and it is therefore useful in the realistic applications of human-computer reciprocations, security and monitoring systems and in the care of mentally ill individuals.