Physical controls for most devices are either "one size fits all" or require custom hardware for each user. Cost often prohibits custom design, and each user must adapt to the standard device interface, typically with a loss of precision and efficiency. When user abilities vary widely, such as in the disabled community, devices often become unusable. The goal is to create a system that will track user gestures and interpret them as control signals for devices. Existing gesture recognition research converts continuous body motion into discrete signals. The approach is to map continuous motions into a set of analog device control signals. The system will allow us to quickly tailor a device interface to each user's best physical range of motion. The first application domain is a speech synthesizer for disabled users. Results are expected in two major areas of applicability for non-disabled users: in telemanipulator interfaces, and as a design tool for creating biomechanically efficient interfaces.
