A Viola Jones Control Model for Emotion Recognition Based on Facial Expression and Body Gestures in Real-time Closed loop
1. Abstract
Body gestures are one of the crucial components of body language and can also be used for emotion recognition in people. Emotional epitome can also be derived from body language and what is going on within the confines of the body language that interprets the emotion in a person’s face. This could include movement and depths and changes in the motion of the body in a confined environment. But these aspects remain less explored to recognise emotion while facial expression and examples such as speech-based approaches are widely investigated. This paper introduces a custom method of emotion recognition through facial expression and body gestures in people based on Viola- Jones using Matlab images saved in archive and a thousand images of body gestures and facial expressions. A Viola Jones control model was used to save these images and to identify features that recognise different forms of emotion saved in the database. The surrounding environment of the person also helps to map the right emotion expression on the face of the person. Tracking points and error tracking was computed for each class of emotion detected; prediction accuracy is close to 78% for each predicted class of emotion.
Keywords
Body gestures, facial expression, Voila Jones, Point tracking, emotion recognition, Speed-based emotion recognition