Merge pull request #3 from jogloran/patch-1
Browse files
readme.md
CHANGED
|
@@ -107,7 +107,7 @@ latents.backward(gradient=grad, retain_graph=True)
|
|
| 107 |
* Shading & normal evaluation: `./nerf/network*.py > NeRFNetwork > forward`. Current implementation harms training and is disabled.
|
| 108 |
* use `--albedo_iters 1000` to enable random shading mode after 1000 steps from albedo, lambertian ,and textureless
|
| 109 |
* light direction: current implementation use a plane light source, instead of a point light source...
|
| 110 |
-
* View-dependent prompting: `./nerf/provider.
|
| 111 |
* ues `--angle_overhead, --angle_front` to set the border. How to better divide front/back/side regions?
|
| 112 |
* Network backbone (`./nerf/network*.py`) can be chosen by the `--backbone` option, but `tcnn` and `vanilla` are not well tested.
|
| 113 |
* the occupancy grid based training acceleration (instant-ngp like) may harm the generation progress, since once a grid cell is marked as empty, rays won't pass it later.
|
|
|
|
| 107 |
* Shading & normal evaluation: `./nerf/network*.py > NeRFNetwork > forward`. Current implementation harms training and is disabled.
|
| 108 |
* use `--albedo_iters 1000` to enable random shading mode after 1000 steps from albedo, lambertian ,and textureless
|
| 109 |
* light direction: current implementation use a plane light source, instead of a point light source...
|
| 110 |
+
* View-dependent prompting: `./nerf/provider.py > get_view_direction`.
|
| 111 |
* ues `--angle_overhead, --angle_front` to set the border. How to better divide front/back/side regions?
|
| 112 |
* Network backbone (`./nerf/network*.py`) can be chosen by the `--backbone` option, but `tcnn` and `vanilla` are not well tested.
|
| 113 |
* the occupancy grid based training acceleration (instant-ngp like) may harm the generation progress, since once a grid cell is marked as empty, rays won't pass it later.
|