Where can I find descriptions of the high level features for AcousticBrainz?


Is there one place that describes these terms ?

I saw this page https://acousticbrainz.org/datasets/accuracy, but I would still like more information on terms such as atonal and what they mean.

Also, I’ve seen it mentioned that there is a feature that lets people create datasets from AcousticBrainz data… Where is the link to this ?

Thanks so much!


High level features are computed based on examples given to a machine-learning process. A description of this process is linked from the page that you found: https://github.com/MTG/essentia/blob/master/FAQ.md#how-it-works

This means that, in the case of tonal/atonal, we selected 144 music excerpts which a (human) listener thought sounded atonal, and 199 which they thought sounded tonal.
This means that the results therefore reflect the songs that went into the process. It could be difficult to describe the terms without giving examples. While we can’t share the audio used for these models because of copyright restrictions, we are hoping to in the future add metadata to this accuracy page (song names, mbids, etc) so that people can find the audio themselves and listen to it.

To use the dataset editor, log into your account and then follow the link from your profile. Thanks for reminding me - we need to also publish a tutorial on how to do this.



I went to the dataset editor. Not sure if I understand it.
Does one need to enter one mbid at a time to make a dataset ?


For now, yes. We’re planning other features that lets you add items by search (for example, adding all tracks by an artist, or a specific track by its name).
You can also create a CSV file which contains the mapping between an mbid and the label you want to give it, and upload it to create the dataset.