Binary Force Myography

Force myography infers gestures from the change in limb volume as muscles contract, sensed through force-resistive sensors pressed against the skin — a simpler, lower-cost alternative to electromyography. This project tested how much could be recovered from a deliberately minimal, binary version of the signal.

The result was a low-cost wearable band that classified a useful set of hand gestures from a handful of thresholded sensors.

Contributions

1. Designed a wearable band housing force-resistive sensors around the forearm.
2. Built a binary thresholding scheme for gesture classification.
3. Evaluated accuracy across a set of common hand gestures.