PianoBART: Symbolic Piano Music Generation and Understanding with Large-Scale Pre-Training

TL;DR

PianoBART is a large-scale pre-trained model based on BART that effectively learns musical structures for both symbolic piano music generation and understanding, achieving high-quality results.

Contribution

It introduces a multi-level object selection pre-training strategy for a unified model handling music generation and understanding.

Findings

Outperforms existing models in generating coherent piano music

Achieves superior accuracy in music understanding tasks

Effectively captures musical semantics during pre-training

Abstract

Learning musical structures and composition patterns is necessary for both music generation and understanding, but current methods do not make uniform use of learned features to generate and comprehend music simultaneously. In this paper, we propose PianoBART, a pre-trained model that uses BART for both symbolic piano music generation and understanding. We devise a multi-level object selection strategy for different pre-training tasks of PianoBART, which can prevent information leakage or loss and enhance learning ability. The musical semantics captured in pre-training are fine-tuned for music generation and understanding tasks. Experiments demonstrate that PianoBART efficiently learns musical patterns and achieves outstanding performance in generating high-quality coherent pieces and comprehending music. Our code and supplementary material are available at https://github.com/RS2002/PianoBart.