FantasticOven2

Hey! Sorry for the late reply. > always showing "Warning: image_path not found for reconstruction" You should parse the folder contains the image for reconstruction when calling `simple_trainer_2dgs.py`. You can...

Hey, can you try to run `nvcc --version` in your conda env and what's the output?

Hi, the "images_4" refers to the directory that contains images downsampled 4 times and it is provided in the [mipnerf-360 dataset](https://jonbarron.info/mipnerf360/). When using your own dataset, you can tune the...

I think it's prefered to run colmap on the original images and then downsample the images. You also need to change the camera intrinsics when downsampling. Here's a [reference](https://github.com/graphdeco-inria/gaussian-splatting?tab=readme-ov-file#processing-your-own-scenes) for...

> @fatbao55 Please check this PR, [graphdeco-inria/diff-gaussian-rasterization#10](https://github.com/graphdeco-inria/diff-gaussian-rasterization/pull/10) For my case, adding "-Xcompiler -fno-gnu-unique" option in submodules/diff-gaussian-rasterization/setup.py: line 29 resolves the illegal memory access error in training. > > ``` >...

Forward pass works on simple 2D experiments for v1.0  

Slant views  (3DGS)  (2DGS)

2D Image fitting is working now. (2DGS)  (3DGS)  Go Bairs! 

> Hello, if I want test 2DGS with `ns-train splatfacto`, is it possible? Hey @ichsan2895 , Thanks for your interests! Currently we are integrating 2DGS into gsplat so `ns-train splatfacto`...

So the 3D reconstruction runs now; here's the examples (2DGS, PSNR: 21)  (3DGS, PSNR: 25)  There are bugs that cause blurriness in 2DGS, most likely in the backward...