Guess the Emotions game - Thesis

This study explored the concept of a hands-free augmented reality (AR) game for children on the autism spectrum, in particular on the improvement of emotion recognition skills. At present, 1 in every 59 children in the US and 40 in every 10 thousand children in China are diagnosed with autism spectrum (AS). These stats have risen in recent years, and as such, much research has been devoted to AS and developing tools to aid individuals on the spectrum. For children, toys play an essential role in their development, and research has been conducted on developing toys and games for children on the AS. Some of this research has already taken into account AR games and emotion recognition, but not so much together and none with consideration of a hands-free concept, meaning that all pre-existing research has limited the motor capability of children while playing these games. Thus, this study focused on hands-free AR, and in contribution to the small amount of research in emotional development, the game focuses on improving emotion recognition. For this study, participants consisted of 6 neurodivergent children. Participants could not all be children on the AS due to the difficulty of forming relations with schools and the limited availability among the students. However, in this case, having 6 neurodivergent children allowed for insight on varying levels of emotion recognition dissimilar to neurotypical children. Apart from surveys to test their skills in emotion recognition before and after playing the game, the game itself was designed for generating monsters with various facial expressions, conveying one of five basic emotions. It was prototyped and used by the participants between the two sessions. The game included both monsters and virtual humans to familiarize the children with different versions of the same expression. Changes, such as sound and the scaffolding rules for emotions and faces, were made to the prototype following the first session, based on the insigh