Evolvable Lip Contour Model for Emotion Recognition

摘要

This paper proposes an alternative approach to emotion recognition from the outer Up contour of the subjects. Subjects exhibit their emotions through their facial expressions, and the lip region is segmented from their facial images. A lip contour model has been developed, and the parameters of the model are adapted using differential evolution algorithm to match with the actual outer contour of the lip. An SVM classifier is then employed to classify the emotion of the subject from the parameter set of the subjects' lip contour. The experiment was performed on 50 subjects in the age group 20-30 years, and the worst accuracy in emotion classification is found to be 86%.