Advancing Multi-Instrument Music Transcription: Results from the 2025 AMT Challenge
Sound
2026-03-31 v1 Information Retrieval
Abstract
This paper presents the results of the 2025 Automatic Music Transcription (AMT) Challenge, an online competition to benchmark progress in multi-instrument transcription. Eight teams submitted valid solutions; two outperformed the baseline MT3 model. The results highlight both advances in transcription accuracy and the remaining difficulties in handling polyphony and timbre variation. We conclude with directions for future challenges: broader genre coverage and stronger emphasis on instrument detection.
Keywords
Cite
@article{arxiv.2603.27528,
title = {Advancing Multi-Instrument Music Transcription: Results from the 2025 AMT Challenge},
author = {Ojas Chaturvedi and Kayshav Bhardwaj and Tanay Gondil and Benjamin Shiue-Hal Chou and Kristen Yeon-Ji Yun and Yung-Hsiang Lu and Yujia Yan and Sungkyun Chang},
journal= {arXiv preprint arXiv:2603.27528},
year = {2026}
}
Comments
7 pages, 3 figures. Accepted to the AI for Music Workshop at NeurIPS 2025