As we are not allowed (?) to store the datasets here, we need to remove them and instead write scripts to download/unzip/ and use them.

1. It can be done through pyqt user interface
2. for SDD/GC at least we might store the annotations separately somewhere as that would be too much to download all the videos