Saved in:
Bibliographic Details
Main Authors: Hao, Chunbo, Yuan, Ruibin, Yao, Jixun, Deng, Qixin, Bai, Xinyi, Wang, Yanbo, Xue, Wei, Xie, Lei
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.02797
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • Music structure analysis (MSA) underpins music understanding and controllable generation, yet progress has been limited by small, inconsistent corpora. We present SongFormer, a scalable framework that learns from heterogeneous supervision. SongFormer (i) fuses short- and long-window self-supervised learning representations to capture both fine-grained and long-range dependencies, and (ii) introduces a learned source embedding to enable training with partial, noisy, and schema-mismatched labels. To support scaling and fair evaluation, we release SongFormDB, the largest MSA corpus to date (over 14k songs spanning languages and genres), and SongFormBench, a 300-song expert-verified benchmark. On SongFormBench, SongFormer sets a new state of the art in strict boundary detection (HR.5F) and achieves the highest functional label accuracy, while remaining computationally efficient; it surpasses strong baselines and Gemini 2.5 Pro on these metrics and remains competitive under relaxed tolerance (HR3F). Code, datasets, and model are open-sourced at https://github.com/ASLP-lab/SongFormer.