About one sample in Manifold40

[MyEternalMoon]

Hi, thanks for you great paper and code! When I want to make use of the new dataset Manifold40 you proposed from tsinghua cloud to training my own model, I found that there was a mere mistake in only one sample that is 'bookshelf_0099.obj'. The content of the version I downloaded is:

Object res.obj

Vertices: 245

Faces: 842

with 842 faces, not 500, so I remeshed it in meshlab to make the code works.

[lzhengning]

Hi @MyEternalMoon ,

Thanks for hightlighting the data sample.

There are a lot of holes / genues in the mesh 'bookshelf_0099'. So I tried my best to make it manifold with the genus unchanged. As a result it has 842 faces rather than 500 faces.

Even so, I cannot further remesh 'bookshelf_0099' to have subdivision connectivity. So I discarded this training sample (as mentioned in the Appendix).

If you properly simplified the number of faces in 'bookshelf_0099' to 500, you are welcomed to share it with me. I will update the Manifold40 dataset.

[MyEternalMoon]

@lzhengning Thanks for your reply:)