This page hosts the dataset files for the paper, which was presented at ISMIR 2023. This was carried out by me, Xavier Riley, a PhD candidate on the AIM programme at QMUL.
This dataset contains audio, aligned midi and MusicXML scores for 48 tracks in the FiloBass dataset. These can be used to train automatic transcription models, beat tracking models or for musicological… See the full description on the dataset page:
https://huggingface.co/datasets/xavriley/FiloBass.
