Fig. 2: Overview of PointNet and 3D Sparse CNN architectures. | npj Computational Materials

Fig. 2: Overview of PointNet and 3D Sparse CNN architectures.

From: Deep learning for symmetry classification using sparse 3D electron density data for inorganic compounds

Fig. 2

a PointNet: Visualization of the PointNet architecture, focusing on the symmetric function. b Sparse 3D CNN Process: This illustration delineates the sparse convolutional process, beginning with rulebook creation and execution of sparse convolution. The visualization of the PointNet and 3D Sparse CNN architectures in Supplementary Fig. 2, and more detailed descriptions of the PointNet and 3D Sparse CNN functioning are presented in Supplementary Fig. 3.

Back to article page