About the details of the code

[ideasplus]

Hello, thanks for your open-source code!

I want to know which part of the code is responsible for identifying occlusion and signal miss? I can't find it due to the complexity of this project. And can you give more explaination on the intution of identifying occlusion and signal miss in the spherical coordinate system? I can't understand it well.

Another question is about the process of model inference. Do we need to assemble the approximated complete shape for the test sample or just take the original sample as input and output the shape occupancy probablity?