The pretrained network most likely won't generalize to 2D point clouds. But from the algorithm point of view, I don't think there's anything stopping you from using 2D point clouds. That said, you should change the models to handle 2D registration, e.g. changing the pose estimation to estimate 2D rotation+translation instead of the given 3D one.
The pretrained network most likely won't generalize to 2D point clouds. But from the algorithm point of view, I don't think there's anything stopping you from using 2D point clouds. That said, you should change the models to handle 2D registration, e.g. changing the pose estimation to estimate 2D rotation+translation instead of the given 3D one.
Zi Jian