Comparison Gestures Recognition Using K-NN and Naïve Bayes

Abstract

There are some technics to recognize the gesture of the hands using electromyography signals (EMG). Comparing the K-NN method and the Naïve Bayes algorithm is presented in this paper. To obtain the EMG signal, a Myo armband is used, which is placed in the arm of a subject. Five gestures of the hand are used to be compared by both algorithms. Those gestures are utilized to drive a mobile robot to follow a particular path. The results show that the Naïve Bayes has success scores higher than the K-NN. Moreover, when the algorithms are used to control the robot, the Naïve Bayes method is faster to complete the path than the K-NN algorithm.