labels of videos to original COVLA dataset

Hi, there. Great work!

I'm wondering if there is a dictionary of categorized videos to the original COVLA dataset? For example, for each video in act-bench, the scene_id in the COVLA dataset. I think that the category is very important to use the COVLA data without being dominated by simple maneuvers (like straight).

Looking forward to your help.
Thank you.