about values of tables in the paper, 32*3, 64*3

[caixialiu-bjut]

Hi, your work inspires me. How does 3D-RecGAN++ get the IoU of 323 voxels, 643 voxels in tables? I think that the output is 2563 voxels by main_3D-RecGAN++.py.
We observe that the channels of 323 voxels, 643 voxels, 1283 voxels are respectively 64, 16, 8 in the architecture of 3D-RecGAN++. Is the average IoU of the channels, i.e., the IoU of 323 voxles is the average IoU of 64 channels of 323 voxels? Thanks very much, look forward to your reply.

[Yang7879]

hi @caixialiu-bjut, if you need to compute the IoU of 32^3 or 64^3, you can downsample the predicted 256^3 voxels to 32^3 or 64^3. The intermediate layers are features, not voxel results.