Thanks for releasing code! I have some question about the ground-truth object mesh in the replica dataset:
1.Line 27-28 in data_generation/extract_inst_obj.py, why does sub_mesh_indices[object_id] append twice?
The ground-truth object mesh has holes due to occlusion. How can we faithfully evaluate the reconstructed object-level mesh on the occluded surface?
For example, there are holes on the sofa due to the pillows on top of it.
Sorry, I can't really remember why. Does it work well without appending twice? If so, it might be a typo. I think the object-level mesh is still correct.
I tried to find the complete mesh at the object level, but this is the best object-level GT mesh I can get from the Replica Dataset by splitting the scene mesh. I didn't find a better dataset that both contains good/diverse object/scene level mesh and supports data rendering. Please let me know if I missed any good datasets.
Hi,
Thanks for releasing code! I have some question about the ground-truth object mesh in the replica dataset:
1.Line 27-28 in data_generation/extract_inst_obj.py, why does
sub_mesh_indices[object_id]
append twice?