Facial expression recognition using computationally intelligent techniques

Abstract

Facial expressions classification is a fast growing research area. Lots of contribution has been made in this area by researchers from fields of computer science, computer vision, artificial intelligence and psychology. There are many applications that use facial expression classification to identify the behavior, emotion, feelings and opinion of a person. Facial expression classification is not a trivial task as there are many factors that need to be accounted like low quality of images, noise, and shape/color of image. In this article, we have proposed an efficient facial expression classification scheme. In the first step, we perform some pre-processing steps like face detection and histogram equalization inorder to reduce the data dimenions and normalize the illumination effects. Then, an efficient feature extraction technique is used to extract the relevant face features. In the last step, we train and test Support Vector Machine (SVM) classifier to classify the facial expressions. Emirical results obtained using the JAFFE database suggest that the proposed technique produces impressive results by utilzing the best facial features.