question

Hi, thanks for sharing this great work!
I see that the current implementation uses the SHHS dataset.
Could this framework also be applied to the Sleep-EDF dataset?
If yes, are there any key changes needed in the preprocessing step?