Japanese syllabary identification using myoelectric potential of neck muscles

The purpose of this paper is to achieve Japanese syllabary identification without using speech signals. For this purpose, I put an array electrode on the anterior surface of neck and measure BEP signals and I propose a method for classifying Japanese syllabary using BEP signals with SVM. As a first step, we conducted an experiment to identify 10 Japanese syllabary (46 kinds) and rest state. As a result, the average of identification accuracy is 96.6% and least 80%.