search

QianyiWu / objsdf

:t-rex: [ECCV‘22] Pytorch implementation of 'Object-Compositional Neural Implicit Surfaces'
183 stars 5 forks source link

The mesh evaluation #12

Closed kyleleey closed 1 year ago

kyleleey commented 1 year ago

Hi Qianyi,

I didn't find the corresponding script of mesh evaluation on Toydesk dataset in your codebase, and also in the objnerf codebase. I wonder how did you get the Chamfer Distance provided in the paper?

Because in the toydesk dataset, it seems that the gt point cloud is with many outliers, I wonder if there is any other operations that need to be done in addition to aligning the output mesh and gt with center and scale.

QianyiWu commented 1 year ago

Hi Kyleleey @kyleleey ,

Sorry for the late reply. I updated the mesh evaluation script for the Toydesk dataset in here. The usage would be python compute_3d.py [path to gt mesh] [path to generated mesh] [center file] In this script, there will be a 3D box to crop the mesh into a valid region. But I failed in finding the exact crop size for the desk 1 and 2... Sorry about that. It might be a little bit different than the current size (1.0) as shown in the script.

Hope this script would be helpful.

kyleleey commented 1 year ago

Hi Qianyi,

No problem, I totally understand this and thanks so much for providing this script