Abstract: Kernel point convolution (KPConv) defines convolutional weights based on Euclidean distances between kernel points and input points and has shown good segmentation results on several ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback