Kerui Ren
Kerui Ren
In fact, the effect of the Scaffold-GS has great improvement compared to the vanilla Gaussian splatting, the latest results are shown in the arxiv version of [Octree-GS](https://arxiv.org/abs/2403.17898).
Please check `train.sh`//16.
In fact, we did not modify the remote viewer for adaptation, and we will consider this issue in the viewer of future work.
Can you view correct result with given examples?
I think this is an interesting idea, and we will think about feature fusion later.
Actually we take this problem into account and add `image_scale` in **LOD Level Selection** for regulation, but the qualitative effect is between Scaffel-GS and Mip-Splatting.
3 of 6 is for scaling the offsets
Refer to the `load_ply` function in `gaussian_model.py` to read the information from the `.ply` file and add the mask.
Please show me your configs.
We update the viewer and you can try it again