Learning Disentangled Representation for Multi-View 3D Object Recognition | IEEE Journals & Magazine | IEEE Xplore