Closed chrenx closed 8 months ago
hi, we only have action label for each sequence, which is reflected in the sequence name. We don't have detailed text annotation for now.
Thanks for your reply! One more thing I wanna double check with you is that the BEHAVE dataset did not offer detailed data on hand poses right? When I visualize the ground truth, it doesn't seem to be quite detailed for the interaction between hands and obj.
yes, we did optimize the hand pose a bit to satisfy multiview constraints, but this does not change much. most poses are very similar to mean hand pose
I wonder if there is any text description paired with each sequence in this dataset since I didn't find any.