Andrea

Don't know if its related, however in my case NeRF produced mostly white images. The issue was that my objects background was white and not transparent. Fixing this solved my...

I'm struggling with camera transformation matrices too. You may find this python-blender script by the authors useful to generate your poses: https://github.com/bmild/nerf/issues/78#issuecomment-710179139

You may find this python-blender script by the authors useful to generate your views: #78 (comment)

To explain myself better, I'd like something like this (standard NeRF output): https://user-images.githubusercontent.com/32450751/141373534-f073f47a-eb3e-4732-8ecb-d28ed448076b.mp4 but I get this (DS-NeRF output): https://user-images.githubusercontent.com/32450751/141373547-eb637e35-fb8e-4ef5-9096-7db42e75aa3a.mp4 Is there a way I can get better results in...

I didn't find any solution in the end

Maybe for this kind of scenario (few inward-looking views of an object) BARF (Bundle Adgusting NeRF) and RegNeRF (Regularizing NeRF) could work better (never tried though).

I substitute these parameters with paths as follows: : /home/jonsnow/svox2/opt/ckpt/my-first-exp/ckpt.npz : /home/jonsnow/svox2/datasets/my-dataset-folder I believe the dataset is used in the rendering process just to understand where the "up" vector of...

I substitute them with paths as follows: : /home/jonsnow/svox2/opt/ckpt/my-first-exp/ckpt.npz : /home/jonsnow/svox2/datasets/my-dataset-folder I believe the dataset is used in the rendering process just to - understand where the "up" vector of...