This project aims to create a pair of motion-detecting gloves that allow the practice of a music instrument without a musical instrument itself. The gloves should be constructed in such a way that each arrangement of finger position and flexion as one would perform on an instrument produces the corresponding sound that would be produced on the instrument. (For example, when in 'clarinet mode,' the flexion of the thumb, index, middle, and ring fingers on the left hand would produce a Bb note, or a transposed C.) The gloves will ideally be constructed using flex-sensors, gyroscopes, and accelerometers, and will not be pre-programmed, but instead have the ability to 'download' new instruments using machine learning.