Facial expression recognition using active shape models and support vector machines

Facial Expression Recognition is the subsequent step after Face Detection and Real time recognition of facial expressions is a challenging task. Various technologies of Facial Expression Recognition has been experimented by researchers over the past few years. In this paper, it has been observed the accuracy and effectiveness of employing Active Shape Models and Support Vector Machines to achieve higher recognition rates. Active Shape Model is used to locate the facial feature deformations of a face detected by using Haar classifiers. These facial coordinates are fed into a Support Vector Machine and the trained system classifies the expressions into seven categories, namely happy, sad, anger, disgust, fear, surprise and neutral. The system was tested on JAFFE Database and Cross Validation had been used as a mechanism for analysing the results of the experiment.