train for ngp datasets failed

[Asxcvbn]

I tried to reconstruct a instant-ngp dataset and got a fairly problematic result shown below. I re-created transforms.json with data/colmap2nerf.py and the -dataset_name ngp method still don't work. However, switching to -dataset_name colmap works. Maybe camera position was misinterpreted in ngp method here

[Asxcvbn]

In instant ngp dataset fox you may notice there are not as many images as its transform.json claims. currently it will fail to run. i'll post a pr just to address this fail-to-run issue. but i dont know how to fix the main issue yet.
pr here #35

[Linyou]

Hi @Asxcvbn. Thanks for reporting this. I have identified a bug in which the wrong camera pose was being used for training. I believe that PR #43 will resolve this issue.