Hi! I really liked you work!
I have a question regarding the second stage of your training pipeline. In particular, it is not clear to me which supervision you use to optimize the FC layer in the 3D generation branch. Do you compute a distance between the input point cloud and the generated one?
Hi! I really liked you work! I have a question regarding the second stage of your training pipeline. In particular, it is not clear to me which supervision you use to optimize the FC layer in the 3D generation branch. Do you compute a distance between the input point cloud and the generated one?
Thanks in advance, Andrea Amaduzzi