PianoMotion10M: Dataset and Benchmark for Hand Motion Generation in Piano Performance

TL;DR

This paper introduces PianoMotion10M, a large annotated dataset of piano hand movements and a benchmark for generating realistic hand motions from audio, aiming to improve piano instruction systems.

Contribution

It provides a comprehensive dataset and evaluation framework for hand motion generation in piano performance, addressing a gap in music instruction AI tools.

Findings

Collected 116 hours of annotated piano videos with 10 million hand poses.

Developed a baseline model that generates hand motions from audio.

Designed evaluation metrics for motion quality and accuracy.

Abstract

Recently, artificial intelligence techniques for education have been received increasing attentions, while it still remains an open problem to design the effective music instrument instructing systems. Although key presses can be directly derived from sheet music, the transitional movements among key presses require more extensive guidance in piano performance. In this work, we construct a piano-hand motion generation benchmark to guide hand movements and fingerings for piano playing. To this end, we collect an annotated dataset, PianoMotion10M, consisting of 116 hours of piano playing videos from a bird's-eye view with 10 million annotated hand poses. We also introduce a powerful baseline model that generates hand motions from piano audios through a position predictor and a position-guided gesture generator. Furthermore, a series of evaluation metrics are designed to assess the performance of the baseline model, including motion similarity, smoothness, positional accuracy of left and right hands, and overall fidelity of movement distribution. Despite that piano key presses with respect to music scores or audios are already accessible, PianoMotion10M aims to provide guidance on piano fingering for instruction purposes. The source code and dataset can be accessed at https://github.com/agnJason/PianoMotion10M.