Difficulty-Controlled Simplification of Piano Scores with Synthetic Data for Inclusive Music Education

2025-11-20

AI Summary

AI-driven piano score difficulty adaptation is hindered by reliance on proprietary datasets and MIDI representations, resulting in poor reproducibility, limited transferability, and insufficient notational readability and performability. This paper introduces the first open-source, end-to-end, difficulty-controllable simplification method for MusicXML—a structured, human-readable score format. We synthesize a difficulty-annotated dataset conditioned on melody and harmony features, eliminating manual labeling. Our approach integrates pretrained difficulty estimation and style identification models within a Transformer-based conditional generation framework. Crucially, it enables fine-grained, interpretable difficulty control directly over structured musical notation. Experiments demonstrate a balanced trade-off among target difficulty accuracy, playability, and musical quality. All code, datasets, and models are publicly released to foster reproducible and extensible research in open music AI.

Technology Category

• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Computer Vision: Diffusion Models for Vision

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Web Mining and Content Analysis: Large pretrained models with web data
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Despite its potential, AI advances in music education are hindered by proprietary systems that limit the democratization of technology in this domain. In particular, AI-driven music difficulty adjustment is especially promising, as simplifying complex pieces can make music education more inclusive and accessible to learners of all ages and contexts. Nevertheless, recent efforts have relied on proprietary datasets, which prevents the research community from reproducing, comparing, or extending the current state of the art. In addition, while these generative methods offer great potential, most of them use the MIDI format, which, unlike others, such as MusicXML, lacks readability and layout information, thereby limiting their practical use for human performers. This work introduces a transformer-based method for adjusting the difficulty of MusicXML piano scores. Unlike previous methods, which rely on annotated datasets, we propose a synthetic dataset composed of pairs of piano scores ordered by estimated difficulty, with each pair comprising a more challenging and easier arrangement of the same piece. We generate these pairs by creating variations conditioned on the same melody and harmony and leverage pretrained models to assess difficulty and style, ensuring appropriate pairing. The experimental results illustrate the validity of the proposed approach, showing accurate control of playability and target difficulty, as highlighted through qualitative and quantitative evaluations. In contrast to previous work, we openly release all resources (code, dataset, and models), ensuring reproducibility while fostering open-source innovation to help bridge the digital divide.

Problem

• Developing AI methods for adjusting piano score difficulty to enhance music education accessibility
• Overcoming reliance on proprietary datasets that hinder research reproducibility and extension
• Addressing MIDI format limitations by using MusicXML for better readability and layout

Innovation

• Transformer-based method adjusts MusicXML piano scores difficulty
• Synthetic dataset pairs scores by estimated difficulty levels
• Openly releases all resources ensuring reproducibility and accessibility