MyoTyper: A MYO-based Texting System for Forearm Amputees

Arm gestures are not only important in hand-busy applications, but also helpful for arm amputees when they are interacting with computers. With the motivation of interacting with computers without hand or arm, we propose MyoTyper, a method enables people to interact with computers by using forearm gesture. We employ acceleration sensors and electromyography sensors to collect signals for arm gesture recognition. A disc-shaped virtual keyboard is designed for fast text input by arm gesture set. Our preliminary study shows that our system prototype achieve classification accuracy with 70% input speed of 2.56 words per minute, and the input speed increases when using proficiency improved.