Tony Louis
Tony Louis
I think it has better gradients than Dice.
so I think what the network has actually learned is a space behind some regular shaped frame. Not a glass.
same error for me, any progress?
If my understanding is right, proprocessing the point cloud before the forward pass is used to speed up the training time. While for the inference time, the total amount of...
Thanks a lot! Are we allowed to change the initial grid size before fitting the knn?