Ming-Ming Xiao

Semantic Segmentation for 3D Point Clouds with Curvature-Aware Sampling and Inverse-Density Weighting

The Curvature-Aware Sampling Network (CASNet) is proposed, a feature-preserving framework that combines curvature-aware sampling, inverse-density weighting, attention pooling, and residual blocks that maintains competitive predictive performance while providing favorable computational efficiency.