Gesture Prediction Model for the Guitar Fingering Problem

Abstract

In this thesis we provide a method for finding the fingering of a music piece on any type of guitar using a hand model. Adapting to the real world conditions by deploying a model of the user's hand, and considering the constraints of the guitar and the music notes is what makes our method more realistic. We have modeled the movements of the user's hand in such a way that the thumb does not play any role, and the movements of the other four fingers are modeled using a set of kinematics equations. We use two sets of constraints derived from the guitar and the music notes in order to include the playing techniques, which are required by the music piece and the guitar. The guitar is considered to be a separate entity in our model having its own properties, resulting in a method independent of the type and tuning of the instrument. Since we are using the hand model for generating the fingering of the music piece, the results of the method are gestures generated for the notes, and the final outcome will be an animation for the entire sheet of music.