Rotation Invariant Convolutions for 3D Point Clouds Analysis
  • Effective end-to-end rotation invariant convolutions are designed for 3D point cloud deep learning
  • SOTA performance is achieved on classification, retireval, segmentation, and normal estimation
  • Project Page

        Efficient 3D Point Cloud Convolutional Neural Networks
  • Efficient and effective end-to-end convolutional neural networks for 3D point cloud analysis
  • SOTA performance is achieved on classification and segmentation
  • Project Page