Abstract:
In the recent past, surface Electromyography (sEMG) signals has been effectively applied in control interfaces
for robots, prosthetic hands, legs and other areas. Few works has focused on neck EMG control in spite of the neck being
central to human movements and engagements. We propose a neck EMG model that estimates head movements and apply
the signals to control a game developed in Unity 3D engine. The game involves randomly falling and accelerating objects.
The user slides the games’ player character (paddle) using neck EMG to intercept as many objects. EMG signal is rectified
and FIR filtered and used to control the position of a sliding probe of the game in real-time. The paper utilizes an equation
model and a Machine Learning model to translate the filtered signal to corresponding neck rotation. From the results, all
the subjects reported accuracy of over 70% which confirm the applicability of the interface and ease of use in that there is
little or no calibration needed. The control model using equation showed intuitive direction and speed control while the
Machine Learning scheme offered more stable direction control but limited speed control of the sliding paddle. The control
interface can be applied in multiple areas that involve neck activity e.g. in rehabilitation as well as gaming interface to
enable entertainment for the disabled.
