Comparison EMG Pattern Recognition Using Bayes and NN Methods

Abstract

This article presents a comparative study for recognition hand gesture between Naïve Bayes and Neural Network (NN) methods for electromyography signals (EMG). EMG signals are obtained from five gestures from a subject. Myo Armband is used to receive the information from the upper hand of the user. The outputs of the systems are used to control a mobile robot. The results showed that the NN method has a higher percentage of success than the Naïve Bayes algorithm. However, the Naïve Bayes method is faster than the NN algorithm.