chamfer distance for 4D point cloud dataset,such as Semantic KITTI. change from the implement Due to JIT compile,you can use the code without installing it.
from chamfer4D import ChamferDistance
chamfer_dist = ChamferDistance()
dist1, dist2 = chamfer_dist(points, points_reconstructed)
loss = (torch.mean(dist1)) + (torch.mean(dist2))