Expected Results
In this paper, we exposed some concepts of a game to stimulate children with down syndrome. Our next steps are conclude the development of the JECRIPE game, specially the movement recognition module mentioned in this text. Our goal is to offer an easy interaction for users with impairment, by recognizing broad gestures in front of a single camera. This kind of interaction may facilitate the interaction of this kind of user in comparison to a keyboard mouse and gamepad.
We are implementing the necessary Computer Vision algorithms for movement recognition in CUDA. We plan to adapt or extend the model-based hand tracking using a hierarchical Bayesian filter presented in for our needs. This extension is targeted to recognize broad body movements, instead of hand tracking only.
Finally, we expect to make experiments with children with Down syndrome in pre-school age (3 to 5 years old) to assure that this interaction paradigm is interesting for this kind of user. We believe that our work promises contributions to the Human-Computer Interaction and computer vision areas, at the same time helping people with special needs by the use of computer games