Hand Input Classification

The majority of on-body interfaces are limited by their ability to resolve inputs at high-resolution proximities. We address this problem for the hand and arm region by using innovative approaches to generate feature sets and classify the input signals. Incorporating this design into modern electronic devices may expedite the transition to, and increase the use of, on-body interfaces. This paper discusses the progress in prototyping and classifying the hand input signals.

This research was undertaken as part of Carnegie Mellon University's Fall 2012 18-797 Machine Learning with Signal Processing course.