However phenomena such as 3D shape processing, surface invariance, occlusion, natural parts and the global advantage were absent in trained networks.
However phenomena such as 3D shape processing, surface invariance, occlusion, natural parts and the global advantage were absent in trained networks.