Meta-Learning for Recalibration of EMG-Based Upper Limb Prostheses

Abstract

An EMG-based upper limb prosthesis relies on a statistical pattern recognition system to map the EMG signal of residual forearm muscles into the appropriate hand movements. As the EMG signal changes each time the user puts the prosthesis on, an efficient method for prosthesis recalibration is needed. Here we show that meta-learning is a promising approach for achieving this aim. Furthermore, we show that meta-leaning can be used to recalibrate the prosthesis even when the examples of some movement types are missing in the target session.

Tasks

Reproductions