A Study of Left Fingering Detection Using CNN for Guitar Learning

For performance techniques of guitar playing, there are fingering of the left hand and picking of the right hand. With the fingering fingers of the left hand, it is important to minimize the movement of the fingers and minimize waste, fingers that hold the strings and how to hold the strings. In picking with the right hand, the dimensions of the stroke, timing of playing strings and picking direction are important. In this paper, we describe a motion detection method for guitar players using image recognition and CNN (Convolutional Neural network) as a machine learning technology, using Kinect V2 and MIDI-Pickup in combination.