Identifying different facial expressions from Kinect

Abstract

This project aims to differentiate human emotions by analyzing their facial expressions. In this project, subjects will be invited to take part in an experiment where their facial expressions will be captured while watching various types of movies. These movies aim to evoke different responses from the participants. Their facial expressions will be recorded by Kinect camera. Next, features will be extracted from the recorded videos and classified via machine learning algorithms. After learning the classifier, the author will identify different types of videos by distinguishing the facial expressions in them.