UniqAI/gaussian-splatting
1
0
Fork 0
mirror of https://github.com/graphdeco-inria/gaussian-splatting synced 2024-11-22 08:18:17 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Clarification

Browse Source
This commit is contained in:
bkerbl 2023-07-12 12:10:03 +02:00
parent 35b89fca85
commit 7e126c02c6
3 changed files with 6 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
README.md
View File

@ -301,12 +301,12 @@ cmake --build build --target install
``` ```
#### Ubuntu 20.04 #### Ubuntu 20.04
Backwards compatibility with Focal Fossa is not fully tested, but building SIBR should still work by first invoking Backwards compatibility with Focal Fossa is not fully tested, but building SIBR should still invoking
```shell ```shell
git checkout fossa_compatibility git checkout fossa_compatibility
git submodule update --init git submodule update --init
``` ```
and then continuing with the steps for Ubuntu 22.04. in ```SIBR_viewers``` and then continuing with the steps for Ubuntu 22.04.
### Navigation in SIBR Viewers ### Navigation in SIBR Viewers
The SIBR interface provides several methods of navigating the scene. By default, you will be started with an FPS navigator, which you can control with ```W, A, S, D, Q, E``` for camera translation and ```I, K, J, L, U, O``` for rotation. Alternatively, you may want to use a Trackball-style navigator (select from the floating menu). You can also snap to a camera from the data set with the ```Snap to``` button or find the closest camera with ```Snap to closest```. The floating menues also allow you to change the navigation speed. You can use the ```Scaling Modifier``` to control the size of the displayed Gaussians, or show the initial point cloud. The SIBR interface provides several methods of navigating the scene. By default, you will be started with an FPS navigator, which you can control with ```W, A, S, D, Q, E``` for camera translation and ```I, K, J, L, U, O``` for rotation. Alternatively, you may want to use a Trackball-style navigator (select from the floating menu). You can also snap to a camera from the data set with the ```Snap to``` button or find the closest camera with ```Snap to closest```. The floating menues also allow you to change the navigation speed. You can use the ```Scaling Modifier``` to control the size of the displayed Gaussians, or show the initial point cloud.

4
full_eval.py
View File

@ -12,7 +12,7 @@
import os import os
from argparse import ArgumentParser from argparse import ArgumentParser
mipnerf360_outdoor_scenes = ["bicycle", "flowers", "garden", "stump", "treehill"] mipnerf360_outdoor_scenes = ["flowers", "garden", "stump", "treehill"]
mipnerf360_indoor_scenes = ["room", "counter", "kitchen", "bonsai"] mipnerf360_indoor_scenes = ["room", "counter", "kitchen", "bonsai"]
tanks_and_temples_scenes = ["truck", "train"] tanks_and_temples_scenes = ["truck", "train"]
deep_blending_scenes = ["drjohnson", "playroom"] deep_blending_scenes = ["drjohnson", "playroom"]
@ -37,7 +37,7 @@ if not args.skip_training or not args.skip_rendering:
args = parser.parse_args() args = parser.parse_args()
if not args.skip_training: if not args.skip_training:
common_args = " --quiet --eval --test_iterations -1" common_args = " --eval --save_iterations -1"
for scene in mipnerf360_outdoor_scenes: for scene in mipnerf360_outdoor_scenes:
source = args.mipnerf360 + "/" + scene source = args.mipnerf360 + "/" + scene
os.system("python train.py -s " + source + " -i images_4 -m " + args.output_path + "/" + scene + common_args) os.system("python train.py -s " + source + " -i images_4 -m " + args.output_path + "/" + scene + common_args)

2
scene/gaussian_model.py
View File

@ -84,7 +84,9 @@ class GaussianModel:
self.active_sh_degree += 1 self.active_sh_degree += 1
def create_from_pcd(self, pcd : BasicPointCloud, spatial_lr_scale : float): def create_from_pcd(self, pcd : BasicPointCloud, spatial_lr_scale : float):
spatial_lr_scale = 5
self.spatial_lr_scale = spatial_lr_scale self.spatial_lr_scale = spatial_lr_scale
#print(spatial_lr_scale)
fused_point_cloud = torch.tensor(np.asarray(pcd.points)).float().cuda() fused_point_cloud = torch.tensor(np.asarray(pcd.points)).float().cuda()
fused_color = RGB2SH(torch.tensor(np.asarray(pcd.colors)).float().cuda()) fused_color = RGB2SH(torch.tensor(np.asarray(pcd.colors)).float().cuda())
features = torch.zeros((fused_color.shape[0], 3, (self.max_sh_degree + 1) ** 2)).float().cuda() features = torch.zeros((fused_color.shape[0], 3, (self.max_sh_degree + 1) ** 2)).float().cuda()