Semantic Segmentation of Point Cloud Scene via Multi-Scale Feature Aggregation and Adaptive Fusion

Point cloud semantic segmentation is a key step in 3D scene under-standing and analysis. In recent years, deep learning-based point cloud semantic segmentation methods have received extensive at-tention from researchers. Multi-scale neighborhood feature learning methods are suitable for inhomogeneous density point clouds, but different scale branching feature learning increases the computational complexity and makes it difficult to accurately fuse different scale fea-tures to express local information. In this study, a point cloud semantic segmentation network based on RandLA-Net with multi-scale local fea-ture aggregation and adaptive fusion is proposed. The designed struc-ture can reduce computational complexity and accurately express local features. The mean intersection-over-union is improved by 1.1% on the SemanticKITTI data set with an inference speed of nine frames per sec-ond, while the mean intersection-over-union is improved by 0.9% on the S3DIS data set, compared with RandLA-Net. We also conduct abla-tion studies to validate the effectiveness of the proposed key structure