A scheme of features fusion for facial expression analysis: A facial action recognition

Abstract

The study of emotions is very tough job which sounds in the form to recognize the facial expressions of human face. This problem is quite challenging because of various face architecture of the facial parts of human. In fact it is variant with locality. It becomes hot topic of research for hearing impaired people. Facial expression is considered as a part of sign language recognition (SLR).The pre-processing involves face detection using the algorithm proposed by Viola and Jones using Haar-cascades for facial part detection. In this work, we purpose a fusion of texture local binary pattern, landmark based active shape model (ASM) of geometric features. The features are trained and tested by multi-class support vector machine (SVM).