AVATAR: Human-Computer Interface for Interaction with Children using a Live Animation Process Based in Facial and Body Landmarks Recognition

Abstract

This document shows the process through which we went through for extracting key points in the pose and devices used as well as in extracting the points from the face, which was used image processing and a classifier. To provide movement to computer-created models based on human models, the Blender tool was used to provide add-ons needed for the proper functioning and adequacy of different detections. These models will deal with social interaction with children who have some kind of difficulty communicating with adults, this interaction will provide a better view on whether our proposed system is reliable or needs some adjustments on the right functioning.