Segment-Factorized Full-Song Generation on Symbolic Piano Music

Ping-Yi Chen, Chih-Pin Tan, Yi-Hsuan Yang
10/7/2025
cs.SDcs.AIcs.LG

No official code found. Best implementation: eri24816/segmented-full-song-gen (11 stars, last commit 2025-11-27).

Abstract

We propose the Segmented Full-Song Model (SFS) for symbolic full-song generation. The model accepts a user-provided song structure and an optional short seed segment that anchors the main idea around which the song is developed. By factorizing a song into segments and generating each one through selective attention to related segments, the model achieves higher quality and efficiency compared to prior work. To demonstrate its suitability for human-AI interaction, we further wrap SFS into a web application that enables users to iteratively co-create music on a piano roll with customizable structures and flexible ordering.

View on arXivView PDF

Code Implementations(1)

eri24816/segmented-full-song-genCommunity65%

NeurIPS 2025 AI4Music Workshop: Segment-Factorized Full-Song Generation on Symbolic Piano Music

110PythonOct 6, 202510 months ago

Similar Papers

Via Score to Performance: Efficient Human-Controllable Long Song Generation with Bar-Level Symbolic Notation

84% match2025

Whole-Song Hierarchical Generation of Symbolic Music Using Cascaded Diffusion Models

83% match2 repos572024

TOMI: Transforming and Organizing Music Ideas for Multi-Track Compositions with Full-Song Structure

81% match1 repo122025

A Traditional Approach to Symbolic Piano Continuation

81% match2025

Motifs, Phrases, and Beyond: The Modelling of Structure in Symbolic Music Generation

81% match2024

PianoBART: Symbolic Piano Music Generation and Understanding with Large-Scale Pre-Training

81% match1 repo232024

Cite this paper

@article{chen2025segmentfactorized,
  title  = {Segment-Factorized Full-Song Generation on Symbolic Piano Music},
  author = {Ping-Yi Chen and Chih-Pin Tan and Yi-Hsuan Yang},
  year   = {2025},
  eprint = {2510.05881},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2510.05881v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2510.05881.svg)](https://codeofpaper.com/paper/2510.05881)

Discussion