Ground truth binvox about 3dmodelingrl

[clinplayer]

Hi, I think your command looks good. I remember we used a similar way to generate the voxelized shapes. What do you mean by "does not match"? Maybe some of our categories are not centralized, whereas I think these differences in scale or alignment do not affect the results too much because the primitives are distributed evenly in the voxel space.

Note in this implementation you also need a low-resolution voxel in 16^3 for Prim-Agent for better efficiency and abstraction.

[stephan000]

Hi,

What do you mean by "does not match"?

It means that the voxel data of the model you provide is different from the model obtained by the above command. The models look similar, so if scale and alignment don't affect the results, it doesn't matter.

Thank you.