sunset comments

Results 72 comments of


                                            sunset

cuda extension error

I encountered segmentation fault when using the wrong pytorch or cuda version. Can you verify your pytorch installation is matched with your cuda version?

Test views have low PSNR for Custom Inward Facing Dataset

I have tested `configs/custom/default_ubd_inward_facing.py` on 300 frames subsampled from my casually captured video. It should work. Some general question: 1. I assume you used LLFF's `imgs2poses.py` to generate the pose....

Test views have low PSNR for Custom Inward Facing Dataset

Is it a forward-facing scene? If it is you should use `spherify=False` (see [this guide](https://sunset1995.github.io/dvgo/tutor_forward_facing.html) for more detail). If it is not a forward-facining scene, I guess the problem is...

Calculate ground truth depth

A simpler way is: ```python n_step = segment_coo( src=(weights * step_id), index=ray_id, out=torch.zeros([N]), reduce='sum') depth = t_min + stepdist * n_step ```

why set inverse_y=True in nsvf data

I will try my best to write a simple guide to prepare own data. In your case, a simple work around is to add a flag for `run.py` to vertically...

why set inverse_y=True in nsvf data

I will try to find time to work on it after CVPR.

the data structure of tankstemple_data maybe wrong, would you please help to check? thanks

Hi, could you please provide more detail? I directly use the bounded T&T organized by [NSVF](https://github.com/facebookresearch/NSVF) and use the unbounded T&T organized by [nerf++](https://github.com/Kai-46/nerfplusplus).

Building CUDA Extension

Hmm strange. My machine is `torch==1.10.1+cu111` too and it works well. Could you please provide more detail about your `nvcc` version and which version of the DVGO do you use?...

Building CUDA Extension

On my side, my `nvcc -V` is ``` nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2020 NVIDIA Corporation Built on Tue_Sep_15_19:10:02_PDT_2020 Cuda compilation tools, release 11.1, V11.1.74 Build cuda_11.1.TC455_06.29069683_0...

Building CUDA Extension

I found a similar issue in [other repos](https://github.com/NVIDIA/apex/issues/1268). It seems that gcc version matters too. My `gcc --version` is ``` gcc (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0 Copyright (C) 2017 Free Software Foundation,...