Facial Virtual Tracking: A System to Mirror Emotions

Abstract

The capacity for social interaction, communication and recognition of emotions is a characteristic that the human being possesses and that allows him/her to be socially included. However, some have difficulties in expressing and interpreting emotional states, which can contribute to their marginalisation in society. A particular case is children with Autism Spectrum Disorder. These children have difficulties in social interaction and manifest repetitive patterns. In this work, a serious game is being developed in which an avatar (ZECA avatar) is able to interact with the child, through challenges and training of certain facial movements that will be validated by the system, facilitating imitation and recognition of emotions (happiness, sadness, anger, surprise and fear). The tests performed in the laboratory environment allowed to conclude the adequacy of the game for promoting emotional states in a friendly way.