Facial Expression Recognition Using Artificial Neural Networks

Analysis and recognition of human facial expressions from images and video forms the basis for understanding image content at a higher semantic level. Expression recognition forms the core task of intelligent systems based on human–computer interaction (HCI). In this paper, we explore the use of Artificial Neural Networks in performing expression recognition. We analyze seven basic types of human expressions – neutral, happy, sad, disgust, anger, surprise and fear. The expression recognition task is divided into two steps – an automatic facial feature extraction step, and a classification step that employs Multi Layer Perceptrons and Radial Basis Function Networks. Simulation results demonstrate an expression classification accuracy of 73% on the JAFFE database (using MLP networks), which is significant for a system whose preliminary feature extraction step is automatic.