# SongFormer: Scaling Music Structure Analysis with Heterogeneous Supervision

arXiv: https://arxiv.org/abs/2510.02797 · HTML: https://mlsanity.com/paper/2510.02797/

## Key facts

- Cited by 16 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 3 Oct 2025.

## Built on by (most cited first, 16 of 16)

- [MOSS-Audio Technical Report](https://mlsanity.com/paper/2606.01802/) (2026)
- [AutoMV: An Automatic Multi-Agent System for Music Video Generation](https://mlsanity.com/paper/2512.12196/) (2025)
- [MusTBench: Benchmarking and Advancing Temporal Grounding in Music LLMs](https://mlsanity.com/paper/2605.29300/) (2026)
- [StepAudio 3 Music Technical Report](https://mlsanity.com/paper/2609.16034/) (2026)
- [Hear: Hierarchically Enhanced Aesthetic Representations For Multidimensional Music Evaluation](https://mlsanity.com/paper/2511.18869/) (2025)
- [A Knowledge-Driven Approach to Music Segmentation, Music Source Separation and Cinematic Audio Source Separation](https://mlsanity.com/paper/2602.21476/) (2026)
- [EDMFormer: Genre-Specific Self-Supervised Learning for Music Structure Segmentation](https://mlsanity.com/paper/2603.08759/) (2026)
- [LaDA-Band: Language Diffusion Models for Vocal-to-Accompaniment Generation](https://mlsanity.com/paper/2604.11052/) (2026)
- [VocalParse: Towards Unified and Scalable Singing Voice Transcription with Large Audio Language Models](https://mlsanity.com/paper/2605.04613/) (2026)
- [AgentMV: A State-Guided Multi-Agent Framework for Budget-Aware Music Video Generation](https://mlsanity.com/paper/2605.10723/) (2026)
- [S2Accompanist: A Semantic-Aware and Structure-Guided Diffusion Model for Music Accompaniment Generation](https://mlsanity.com/paper/2605.17414/) (2026)
- [VocalRender: Score-Native Singing Voice Synthesis for Real-World Composition](https://mlsanity.com/paper/2607.27768/) (2026)
- [TimeCues Studio: A Workspace for Music Annotation and Algorithm Prototyping](https://mlsanity.com/paper/2609.10338/) (2026)
- [Open-Qwen-Music: An Auditable Framework for LLM-Based Music Composition and Diffusion Rendering](https://mlsanity.com/paper/2609.31652/) (2026)
- [YuE2: Unifying Symbolic and Audio Music Generation at Frontier Quality](https://mlsanity.com/paper/2609.33757/) (2026)
- [JSIR: An Audio-Aligned, Multi-Task Dataset of Jazz Standard Recordings](https://mlsanity.com/paper/2610.15356/) (2026)

## Abstract (arXiv)

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.