Piano Fingering with Reinforcement Learning

TL;DR

This paper introduces a novel approach to automatic piano fingering by combining knowledge-driven and data-driven methods using deep reinforcement learning, aiming to improve finger placement strategies.

Contribution

It presents a new method that integrates prior knowledge and data-driven learning with reinforcement learning for automatic piano fingering.

Findings

Combines knowledge-driven and data-driven approaches.

Uses deep reinforcement learning for fingering.

Lays groundwork for incorporating past experience.

Abstract

Hand and finger movements are a mainstay of piano technique. Automatic Fingering from symbolic music data allows us to simulate finger and hand movements. Previous proposals achieve automatic piano fingering based on knowledge-driven or data-driven techniques. We combine both approaches with deep reinforcement learning techniques to derive piano fingering. Finally, we explore how to incorporate past experience into reinforcement learning-based piano fingering in further work.