MSGCN: a multiscale spatio graph convolution network for 3D point clouds

We propose a multiscale spatio graph neural network (MSGCN) for 3D point cloud. The core of MSGCN is a multiscale spatio graph(MSG) that explicitly models the relations at various spatial scales. Different from many previous hierarchical structures, the MSG is built in a data adaptive fashion. MSG supports multiscale analysis of point clouds in the scale space and can obtain the dimensional features of point cloud data at different scales. Because traditional convolutional neural networks are not applicable to graph data with irregular vertex neighborhoods, this paper presents an sef-adaptive graph convolution kernel that uses the Chebyshev polynomial to fit an irregular convolution filter based on the theory of optimal approximation. In experiments conducted on four widely used public datasets, The results show that the proposed model outperforms most state-of-the-art methods.