Hoanh Le

Add option to use mask_path while running COLMAP

2

Add option to mask regions while running feature extraction in COLMAP (https://colmap.github.io/faq.html#mask-image-regions)

conflict between scene_box aabb with disable_scene_contraction in instant-ngp / OccGridEstimator

4

`OccGridEstimator` uses different levels of `aabbs` to know where to sample the points from with difference levels of details ``` # multiple levels of aabbs aabbs = torch.stack( [_enlarge_aabb(roi_aabb, 2**i)...

instant-ngp fails to predict rgb when the scene box is too tight?

1

**Describe the bug** I’m trying to limit the region of interest for `instant-ngp` as much as possible to avoid unnecessary computation. However, when I make it very tight around the...

add a way to orient based on main direction of the surface

3

essentially implement https://github.com/nerfstudio-project/nerfstudio/issues/1887 toy problem visualization  `orientation_method == "up" `[Screencast from 08-26-2024 01:31:40 PM.webm](https://github.com/user-attachments/assets/b24d6091-16cf-4b0f-869d-96e3f8fd3b93) `orientation_method == "align" `[Screencast from 08-26-2024 12:53:25 PM.webm](https://github.com/user-attachments/assets/2afee1e4-79e0-4fc2-a1eb-56a00acead89)

c2w incorrect for test images in instant-ngp-data data parser?

1

**Describe the bug** When running ``` ns-train nerfacto --output-dir /home/leh19/workspace/experiment/outputs/base_line --vis viewer+tensorboard --experiment-name painting_1 --pipeline.model.background_color "random" instant-ngp-data --data /home/leh19/test_run_1/JPG/painting_1/processed ``` I got the shape mismatching issue. ``` assert c2w.shape ==...


weird moire pattern when training instant-ngp without scene contraction

1

**Describe the bug** Whenever i trained instant-ngp without scene contraction i often get weird moire pattern on the render view (in both accumulation) and (rgb) **To Reproduce** Train instant-ngp without...

Bad loss spike when training nerfacto with high number of iterations

1

**Describe the bug** I'm training nerfacto on my dataset with 300k iterations (since I want to keep the original data resolution and it's 10 images 6kx4k). I know the standard...