GP-Nerf
GP-Nerf copied to clipboard
The training and testing human id are not consistent with NHP.
Hi,
It seems that the human split in ZjumocapDataset.py
is not in line with the one used in NHP?
For example, the human used for training in GP-NeRF is 313, 377, 394, 315, 386, 390
, while NHP uses 313, 377, 315, 386, 390, 392, 396
for training.
Best,