Tab4U - Automatic Guitar Tablature Generation with Finger Positions using AMT

Abstract

"The guitar is a popular instrument worldwide. However, learning it can take time, especially for those who struggle to find notes without a 'musical ear.' Relying solely on instructions can be frustrating, leading many new players to give up. Guitar tablature, the standard notation, usually doesn't include finger information, which is a challenge addressed in this research. Automatic music transcription (AMT) helps convert audio into a readable format using programming. While some systems transcribe guitar tablature and provide finger information separately, few offer both functionalities together. Integrating these features could significantly assist guitarists in learning new compositions more effectively. This study seeks to address gaps in existing research and products by developing a modular software solution that can work with different AMT algorithms. This thesis outlines the steps involved in developing and evaluating a prototype, showcasing the feasibility of the proposed concept. The creation of a web-based application aims to address the subjective aspects of guitar tablature. Through qualitative assessments, promising findings indicate the potential of the application to enhance guitar learning by providing detailed tablature with finger placement guidance."