Hi,
You don't seem to utilize the 'shared_transform' feature during evaluation; instead, you opt to perform this transform per clip.
I wondered why, as during the, all the transforms should be the same. Therefore, why would you apply it per clip instead of the entire video?
Best,
Orr
Hi, You don't seem to utilize the 'shared_transform' feature during evaluation; instead, you opt to perform this transform per clip. I wondered why, as during the, all the transforms should be the same. Therefore, why would you apply it per clip instead of the entire video? Best, Orr