Abstract
In a previous paper, the authors outlined a technique for motion communication involving functional electrical stimulation in combination with estimation of the user's intended motion. As this approach does not allow information on dynamic motion to be conveyed, however, the authors here outline a technique for the communication of such information between individuals using a state transition model in combination with electrical stimulation and classification of patterns. The method's incorporation of a electrical stimulus-joint angle model also enables the communication of patterns observed in muscle contraction. Experiments performed to verify the applicability of the approach indicated favorable discrimination rates and successful transmission of information on dynamic motion between users, thereby confirming the system's validity.
