Unified Cross-modal Translation of Score Images, Symbolic Music, and Performance Audio

TL;DR

This paper introduces a unified Transformer-based model trained on a large-scale dataset to perform multiple cross-modal music translation tasks, significantly improving accuracy and enabling novel score-image-conditioned audio generation.

Contribution

The paper presents a new large-scale dataset and a unified tokenization framework, allowing a single model to handle various music modality translations simultaneously.

Findings

Reduced optical music recognition error rate from 24.58% to 13.67%.

Achieved first score-image-conditioned audio generation.

Unified model outperforms single-task baselines across multiple translation tasks.

Abstract

Music exists in various modalities, such as score images, symbolic scores, MIDI, and audio. Translations between each modality are established as core tasks of music information retrieval, such as automatic music transcription (audio-to-MIDI) and optical music recognition (score image to symbolic score). However, most past work on multimodal translation trains specialized models on individual translation tasks. In this paper, we propose a unified approach, where we train a general-purpose model on many translation tasks simultaneously. Two key factors make this unified approach viable: a new large-scale dataset and the tokenization of each modality. Firstly, we propose a new dataset that consists of more than 1,300 hours of paired audio-score image data collected from YouTube videos, which is an order of magnitude larger than any existing music modal translation datasets. Secondly, our unified tokenization framework discretizes score images, audio, MIDI, and MusicXML into a sequence of tokens, enabling a single encoder-decoder Transformer to tackle multiple cross-modal translation as one coherent sequence-to-sequence task. Experimental results confirm that our unified multitask model improves upon single-task baselines in several key areas, notably reducing the symbol error rate for optical music recognition from 24.58% to a state-of-the-art 13.67%, while similarly substantial improvements are observed across the other translation tasks. Notably, our approach achieves the first successful score-image-conditioned audio generation, marking a significant breakthrough in cross-modal music generation.