3DCoMPaT$^{++}$: An improved Large-scale 3D Vision Dataset for Compositional Recognition

Open in new window