Open mikatammi opened 8 years ago
Different features need different columns in datasets. Different datasets provide different data. Create some kind of labels that describe what kind of data is available in datasets.
In gazelib/common and with the tools of gazelib, this is solved by having a known set of labels for data streams. See for example https://github.com/infant-cognition-tampere/gazelib/wiki/gazelib-common-v1
However, I have not decided the labels yet.
We should collect the labels in the same place. I suggest we have a gazelib wiki page for that.
Append metadata to dataset to tell what kind of data it provides. This is needed so feature extraction framework knows what features can be extracted from available data.