ProGress: Structured Music Generation via Graph Diffusion and Hierarchical Music Analysis

2025-10-11

AI Summary

Current AI music generation models suffer from weak harmonic-melodic structural coherence and limited musical interpretability. To address this, we propose a structured generative framework integrating Schenkerian analysis with graph diffusion modeling. First, we design a phrase-fusion mechanism grounded in Schenkerian theory to explicitly model tonal hierarchical structure. Second, we extend the discrete graph diffusion model (DiGress) by incorporating hierarchical symbolic music representations and theory-informed probabilistic modeling. Third, we introduce a user-controllable generation interface enabling structural-level interventions—such as harmonic progressions and tonal planning. Listening evaluations demonstrate that our method significantly outperforms state-of-the-art models in structural coherence, musicality, and controllability. To our knowledge, this is the first AI composition paradigm that simultaneously achieves theoretical interpretability, structural editability, and high-fidelity generation.

Technology Category

• Humans and AI: Game Design — Procedural Content Generation & Storytelling
• Machine Learning: Structured Learning
• Natural Language Processing: Generation

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

Artificial Intelligence (AI) for music generation is undergoing rapid developments, with recent symbolic models leveraging sophisticated deep learning and diffusion model algorithms. One drawback with existing models is that they lack structural cohesion, particularly on harmonic-melodic structure. Furthermore, such existing models are largely "black-box" in nature and are not musically interpretable. This paper addresses these limitations via a novel generative music framework that incorporates concepts of Schenkerian analysis (SchA) in concert with a diffusion modeling framework. This framework, which we call ProGress (Prolongation-enhanced DiGress), adapts state-of-the-art deep models for discrete diffusion (in particular, the DiGress model of Vignac et al., 2023) for interpretable and structured music generation. Concretely, our contributions include 1) novel adaptations of the DiGress model for music generation, 2) a novel SchA-inspired phrase fusion methodology, and 3) a framework allowing users to control various aspects of the generation process to create coherent musical compositions. Results from human experiments suggest superior performance to existing state-of-the-art methods.

Problem

• Addresses lack of structural cohesion in music generation
• Solves black-box nature of existing AI music models
• Enhances interpretability and user control in composition

Innovation

• Adapted DiGress model for structured music generation
• Introduced Schenkerian analysis-inspired phrase fusion method
• Developed user-controllable framework for coherent compositions