Efficient Long-Sequence Diffusion Modeling for Symbolic Music Generation

2026-02-28

AI Summary

This work addresses the high computational cost and the dual challenge of modeling long-range dependencies while preserving fine-grained local details in symbolic music generation. To this end, the authors propose SMDIM, a novel diffusion strategy that integrates structured state space models with a selective local refinement mechanism for the first time. This approach enables efficient construction of global musical structure at near-linear complexity while simultaneously performing lightweight optimization of local details. Experimental results demonstrate that SMDIM significantly outperforms existing methods across multiple symbolic music datasets spanning diverse genres, achieving notable advances in generation quality, inference efficiency, and generalization to unseen musical styles.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Generation

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications

Abstract

Symbolic music generation is a challenging task in multimedia generation, involving long sequences with hierarchical temporal structures, long-range dependencies, and fine-grained local details. Though recent diffusion-based models produce high quality generations, they tend to suffer from high training and inference costs with long symbolic sequences due to iterative denoising and sequence-length-related costs. To deal with such problem, we put forth a diffusing strategy named SMDIM to combine efficient global structure construction and light local refinement. SMDIM uses structured state space models to capture long range musical context at near linear cost, and selectively refines local musical details via a hybrid refinement scheme. Experiments performed on a wide range of symbolic music datasets which encompass various Western classical music, popular music and traditional folk music show that the SMDIM model outperforms the other state-of-the-art approaches on both the generation quality and the computational efficiency, and it has robust generalization to underexplored musical styles. These results show that SMDIM offers a principled solution for long-sequence symbolic music generation, including associated attributes that accompany the sequences. We provide a project webpage with audio examples and supplementary materials at https://3328702107.github.io/smdim-music/.

Problem

• symbolic music generation
• long-sequence modeling
• diffusion models
• computational efficiency
• long-range dependencies

Innovation

• diffusion model
• symbolic music generation
• state space model
• long-sequence modeling
• computational efficiency