Anttwo / SuGaR

[CVPR 2024] Official PyTorch implementation of SuGaR: Surface-Aligned Gaussian Splatting for Efficient 3D Mesh Reconstruction and High-Quality Mesh Rendering
https://anttwo.github.io/sugar/
Other
2.34k stars 179 forks source link

Why does 'density' or 'sdf' adapt to the different scenarios? #210

Open wangcong03-shizhuang opened 2 months ago

wangcong03-shizhuang commented 2 months ago

README says that

The density regularization is the simplest one and works well with objects centered in the scene. The SDF provides a stronger regularization, especially in background regions.

What causes this phenomenon? Is this a theoretical result or a practical one?