mir-dataset-loaders / mirdata

Python library for working with Music Information Retrieval datasets
https://mirdata.readthedocs.io/en/stable/
BSD 3-Clause "New" or "Revised" License
351 stars 59 forks source link

IRMAS instrument tagging is not parsed correctly #609

Open GCPhillips opened 8 months ago

GCPhillips commented 8 months ago

Some of the IRMAS training files do not contain a genre tag at the end of the file name, and instead is an instrument tag. For example, if the training track with id 0001__1 is loaded, the genre will be cla and instruments will only contain cel, even though both are instruments and there is no genre tag:

image

In the code, a check could be performed to see if the parsed out string inside the brackets is in INST_DICT.