photometric_confidence & IndexError: min(): Expected reduction dim 0 to have non-zero size.

[iszhihao]

Hello, I use other pre-trained MVSNet models to generate depth maps and confidence values. However, the confidence values generated on the NerfSynth and TankandTemple datasets are all the same, as shown below. photometric_confidence tensor([[[0.1250, 0.1250, 0.1250, ..., 0.1250, 0.1250, 0.1250], [0.1250, 0.1250, 0.1250, ..., 0.1250, 0.1250, 0.1250], [0.1250, 0.1250, 0.1250, ..., 0.1250, 0.1250, 0.1250], ..., Do you know how to resolve this issue? Here is the error message I received: xyz_world_all torch.Size([0, 3]) torch.Size([0, 1]) torch.Size([0]) %%%%%%%%%%%%% getattr(dataset, spacemin, None) None vishull_mask torch.Size([0]) alpha masking xyz_world_all torch.Size([0, 3]) torch.Size([0, 1]) Traceback (most recent call last): File "/cluster/hebut/PointNeRF-v5/run/train_ft_nonstop.py", line 1109, in main() File "/cluster/hebut/PointNeRF-v5/run/train_ft_nonstop.py", line 653, in main points_xyz_all, points_embedding_all, points_color_all, points_dir_all, points_conf_all, img_lst, c2ws_lst, w2cs_lst, intrinsics_all, HDWD_lst = gen_points_filter_embeddings(train_dataset, visualizer, opt) File "/cluster/hebut/PointNeRF-v5/run/train_ft_nonstop.py", line 147, in gen_points_filter_embeddings xyz_world_all, sparse_grid_idx, sampled_pnt_idx = mvs_utils.construct_vox_points_closest(xyz_world_all.cuda() if len(xyz_world_all) < 99999999 else xyz_world_all[::(len(xyz_world_all)//99999999+1),...].cuda(), opt.vox_res) File "/cluster/hebut/PointNeRF-v5/run/../models/mvs/mvs_utils.py", line 541, in construct_vox_points_closest xyz_min, xyz_max = torch.min(xyz, dim=-2)[0], torch.max(xyz, dim=-2)[0] IndexError: min(): Expected reduction dim 0 to have non-zero size. end loading