Feature Extraction and Facial Expression Recognition using Support Vector Machine

Abstract

Facial expressions assume an important part in our everyday collaborations, and late generation has seen an awesome measure of exploring methods for dependable facial impressions identification frameworks. Different innovations of Facial Expression Recognition have been tested by analysts in the course of recent years. Changes in facial expression turn into a troublesome undertaking in perceiving faces. In this we dissect regional facial transformations and utilize various straightforward attributes to shape a compelling classifier. Finally, here exhibited an approach which utilizing an Active Appearance Model and Support Vector Machines. Active Appearance Model (AAM) is used to pull out the unique facial key points and also to consolidate their regional structure attributes to design a classifier. After extracting facial features, these facial coordinates are fed into a Support Vector Machine and the prepared framework classifies the expressions into six classifications specifically like Anger, Fear, Normal, S ad, Disgust and Happy. This framework accomplishes robust and superior expression classification which shows improved results than the existing methods by leading experiments.