Towards Minimum Latency Real-Time Piano Transcription

---

This repository contains the code to run inference and evaluation for the causal minimum-latency piano transcription model described in:

Exploring System Adaptations for Minimum Latency Real-Time Piano Transcription

Authors. Patricia Hu, Silvan Peter, Jan Schlüter, Gerhard Widmer

Presented at ISMIR 2025

Overview

This repository includes inference code and checkpoint to run and evaluate our final model configuration (see Section 4.4 in the paper).

Requirements

Create a new conda environment with Python 3.9 or higher, activate it and run: pip install -r requirements.txt. To run and evaluate the model, run: python src/inference.py path/to/maestro-v3 where path/to/maestro-v3 points to the local directory containing the MAESTRO dataset version 3.

Publication

If you want to find out more, check out our paper.

Acknowledgments

This work is supported by the European Research Council (ERC) under the EU’s Horizon 2020 research & innovation programme, grant agreement No. 10101937 (”Wither Music?”).