Automatic Detection of Social Signals in Digital Playgrounds

Play is a vital activity in which children observe the world, learn new concepts and experiment with them. Even though the social aspect of play is very important, the computer science community has struggled to address it. Digital playgrounds have been built in which children can play in technologically enhanced installations, but the detailed study of the social component of play within these installations has been overlooked. The research proposed here aims to fill in the gap between the human sciences and computer sciences in this context. The emerging field of research called Social Signal Processing (SSP) tasks itself with a goal similar to ours, design systems that are able to detect, interpret and/or reproduce social signals. We plan on using concepts and practices from both the SSP and the Computer Vision fields to analyze the social behavior of children during play in digital playgrounds. We aim to design playgrounds that are able to automatically interpret social interactions and change their dynamics depending on the behavior being exhibited by the children within.