IndexError

[always-anything]

Thank you for your great work! I'm trying your pipeline to reconstrcut the dtu dataset, but failed. Could you please help me to solve this? I have changed a few datasets but they all came into this problem.
Running..
/home/star/software/anaconda3/envs/fds/lib/python3.9/site-packages/torch/init.py:690: UserWarning: torch.set_default_tensor_type() is deprecated as of PyTorch 2.1, please use torch.set_default_dtype() and torch.set_default_device() as alternatives. (Triggered internally at ../torch/csrc/tensor/python_tensor.cpp:451.)
_C._set_default_tensor_type(t)
Load data: Begin
49
Load data: End
total points: 996943
kept points: 69092
Estimating areas for 69092 points
building octree
initializing octree
/home/star/software/anaconda3/envs/fds/lib/python3.9/site-packages/torch/nn/utils/weight_norm.py:28: UserWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.
warnings.warn("torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.")
0%| | 0/20000 [00:00<?, ?it/s]Validate: iter: 1, camera: 44
/home/star/software/anaconda3/envs/fds/lib/python3.9/site-packages/torch/functional.py:507: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at ../aten/src/ATen/native/TensorShape.cpp:3549.)
return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined]
threshold: 0.0
[export.py:78 - export_mesh() ] Exporting 938764 faces as PLY
[exp_runner.py:509 - validate_mesh() ] End
0%|▎ | 46/20000 [00:16<2:00:23, 2.76it/s]
Traceback (most recent call last):
File "/home/star/xzy/fast_dipole_sums/exp_runner.py", line 584, in
runner.train()
File "/home/star/xzy/fast_dipole_sums/exp_runner.py", line 173, in train
data = self.dataset.gen_random_rays_at(image_idx, self.batch_size)
File "/home/star/xzy/fast_dipole_sums/models/dataset.py", line 121, in gen_random_rays_at
p = torch.matmul(self.intrinsics_all_inv[img_idx, None, :3, :3], p[:, :, None]).squeeze() # batch_size, 3
IndexError: index 48 is out of bounds for dimension 0 with size 48

[hanyucc]

Hmm this is strange. Which scene have you tried running on? I just ran the code on a couple DTU scenes and it seems to be working fine. The length of self.intrinsics_all_inv should be equal to 64, which is the number of images in each DTU scene, but it seems like it's 48 in your error message

[always-anything]

Hello and thank you for your attention! I just redownload the DTU dataset again from the link you provided and find out that part of DTU dataset only have 49 images while others have 64. However, when I tried my costom data, the same error happens. Is there any parameters needs to change from different datasets? Or can I just simply change the self.intrinsics_all_inv to ensure this parameter keeps the same with the number of images? Thank you again for your reply!

[hanyucc]

I see. Indeed half of the scenes have 49 images and half of them have 64 images. However, the number of items cameras_sphere.npz contains should be consistent with the number of images. I haven't been able to reproduce this issue: I tried running on scan 24 (which has 49 images) and scan 110 (which has 64 images) and they seem to be working fine. Which scenes are you having issues with?