search

paulguerrero / pcpnet

Pytorch implementation of PCPNet
Other
169 stars 41 forks source link

For multi_scale #4

Closed MulinYu closed 5 years ago

MulinYu commented 5 years ago

Hello, Thanks for ur reponse and sharing.

I noticed that you mentioned in the paper that you translated the patch to the origin and normalize its radius multiplying with 1/r. And you transform output curvatures to the original scale of the point cloud by multiplying with r. But in case of multi_scale, there are more than one r. How do you transform the output curvatures to the original scale?

Have a good day,

best, mulin

paulguerrero commented 5 years ago

Hi Mulin, since there is a constant relationship between the patch sizes, the network can learn to adjust the scale itself when it combines the information from the multi-scale patches.