I have voxels with their occupancy values, i.e. numpy matrix of 0/1 in XYZ grid. Do I have to train the Occupancy Network for one voxel grid, or can I infer it as mentioned in the paper using MultiSurface IsoExtraction with the code and pre-trained model provided? To infer, how should I proceed, and do I have to download the entire dataset of 73GB?
I have voxels with their occupancy values, i.e. numpy matrix of 0/1 in XYZ grid. Do I have to train the Occupancy Network for one voxel grid, or can I infer it as mentioned in the paper using MultiSurface IsoExtraction with the code and pre-trained model provided? To infer, how should I proceed, and do I have to download the entire dataset of 73GB?