Title: MAGI System – Mind Augmented Gesture Interaction
Abstract: MAGI is a research project, part of a collaboration between the MISG group at EIA-FR and the DIVA group at the UNIFR. Gestures are a natural way of interaction. They can be used as a substitute or a complement to other forms of communication. Two key components in gesture recognition are « detection » and « gesture segmentation ». « Detection » refers to the identification of the actor who performs the gestures and the elements involved (hands, head, etc.). « Gesture segmentation » refers to the task of distinguishing a gesture from another and also from the ones not addressed to the machine (involuntary movements or part of the normal gesticulation). In this context, the main idea of the project MAGI is to analyze the physiological sensors’ capabilities – in particular related to the field of Brain Computer Interface (BCI) – in order to enhance the gesture recognition. Specifically, the project has two objectives: on one hand, to enhance the gesture detection by combining the results from conventional sensors (cameras, accelerometers, etc.) with those from physiological sensors (electroencephalogram, electromyogram, etc.); on the other hand, to provide a new approach to solve the problem of gesture segmentation (gestures will be distinguished from each other based on data produced by the user’s brain).
Moreover, a multimodal approach that aims at navigating in a Virtual Word by the use of the only Head (movements, expressions and thoughts) is under analysis and a prototype is in development.