-
Discretizing Group-Convolutional Neural Networks for 3D Geometry in Feature...
A paper that describes how to speed-up deep learning on 3D point clouds with (G-)CNNs by compressing its linear representation layers by a local partial symmetry analysis.