Difficulty-Aware Score Generation for Piano Sight-Reading

2025-09-21

AI Summary

To address the inefficiency of manually designing pedagogically appropriate sight-reading exercises for piano instruction, this paper proposes a difficulty-controllable symbolic music generation method. Methodologically, we introduce a conditional generative framework incorporating control tokens and a lightweight auxiliary difficulty prediction network; difficulty labels serve as explicit constraints, while a difficulty prediction loss acts as an auxiliary optimization objective to mitigate control collapse and ensure precise global difficulty alignment. Experimental results—validated by both automated metrics and expert evaluations—demonstrate that the generated scores exhibit strong difficulty controllability, uniform difficulty distribution, and high pedagogical relevance. To our knowledge, this is the first work to deeply integrate difficulty-aware modeling into the symbolic music generation pipeline, achieving simultaneous preservation of musical quality and substantial improvements in the efficiency and educational adaptability of personalized practice material generation—establishing a scalable technical paradigm for AI-enhanced music education.

Technology Category

• Natural Language Processing: Generation
• Machine Learning: Deep Generative Models & Autoencoders
• Humans and AI: Game Design — Procedural Content Generation & Storytelling

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Adapting learning materials to the level of skill of a student is important in education. In the context of music training, one essential ability is sight-reading -- playing unfamiliar scores at first sight -- which benefits from progressive and level-appropriate practice. However, creating exercises at the appropriate level of difficulty demands significant time and effort. We address this challenge as a controlled symbolic music generation task that aims to produce piano scores with a desired difficulty level. Controlling symbolic generation through conditioning is commonly done using control tokens, but these do not always have a clear impact on global properties, such as difficulty. To improve conditioning, we introduce an auxiliary optimization target for difficulty prediction that helps prevent conditioning collapse -- a common issue in which models ignore control signals in the absence of explicit supervision. This auxiliary objective helps the model to learn internal representations aligned with the target difficulty, enabling more precise and adaptive score generation. Evaluation with automatic metrics and expert judgments shows better control of difficulty and potential educational value. Our approach represents a step toward personalized music education through the generation of difficulty-aware practice material.

Problem

• Generating piano scores with specific difficulty levels for sight-reading practice
• Preventing conditioning collapse in symbolic music generation models
• Creating personalized educational materials that adapt to student skill levels

Innovation

• Auxiliary difficulty prediction objective prevents conditioning collapse
• Learns internal representations aligned with target difficulty levels
• Enables precise difficulty-controlled symbolic piano score generation