Facial Expression Recognition of Animated Human Characters

Abstract

Recognition of animated facial expression is an important and a fundamental task for automatically implementing analysis and evaluation of animation movie products. In this work, an improved algorithm was proposed for recognizing facial expression of animated human characters based on the Local Binary Pattern (LBP) and Support Vector Machine (SVM). The proposed method was tested on an animated facial expression database which contained human characters from top rated animations. The experimental results showed that the recognition accuracy of facial expression was significantly improved compared with other methods.