Toys++ AR Embodied Agents as Tools to Learn by Building

This paper presents a prototype for an augmented-reality based toy. Toys++ is grounded on the concept that the actual activity of building tangible artifacts can speed up learning processes. Toys ++ aims at assembling a framework that will allow the use of existing physical components of the toy as triggers. When the toy is placed under the webcam, a pre-trained 3D feature recognition system scans the entire figure, trying to identify some specific components. If any of these elements are recognized, the system will retrieve and show educational content from selected sources (texts, videos, pictures).