nerfplusplus icon indicating copy to clipboard operation
nerfplusplus copied to clipboard

Code for initializing NeRF's geometry as a sphere

Open yd-yin opened this issue 3 years ago • 3 comments

Hi! Thanks for the elaborate analysis and informative experiments!

Noticed that you conducted an experiment where the geometry (opacity field) of NeRF is fixed (or initialized). Could you please also offer the code of that experiment?

Thanks in advance!

yd-yin avatar Apr 06 '21 04:04 yd-yin

Sorry, I'm not sure if I can find the scripts. But training a opacity field to a unit sphere seems very straightforward? You can simply train the MLP using a loss to ask it to predict 1 on the sphere (x^2+y^2+z^2=1), and zero elsewhere.

Kai-46 avatar Jun 22 '21 21:06 Kai-46

你好!感谢您的精心分析和翔实的实验!

注意到您进行了一个实验,其中 NeRF 的几何形状(不透明度场)是固定的(或初始化的)。您能否也提供该实验的代码?

提前致谢!

Hello, have you solved this problem? How to initialize a 3D model into a sphere?

joeyw1030 avatar Jul 13 '22 08:07 joeyw1030

Refer to the geometric network initialization method from this paper:

SAL: Sign Agnostic Learning of Shapes from Raw Data

csrqli avatar Apr 02 '23 20:04 csrqli