Pointwise Rotation-Invariant Network with Adaptive Sampling and 3D Spherical Voxel Convolution

Published in AAAI, 2020

Yang You, Yujing Lou, Qi Liu, Yu-Wing Tai, Lizhuang Ma, Cewu Lu, Weiming Wang

arXiv / PDF / code / project page / BibTeX

Abstract

In this paper, we propose a new point-set learning framework named Pointwise Rotation-Invariant Network (PRIN), focusing on achieving rotation-invariance in point clouds. We construct spherical signals by Density-Aware Adaptive Sampling (DAAS) from sparse points and employ Spherical Voxel Convolution (SVC) to extract rotation-invariant features for each point. Our network can be applied to applications ranging from object classification, part segmentation, to 3D feature matching and label alignment.

BibTeX

@inproceedings{you2020pointwise,
  title     = {Pointwise Rotation-Invariant Network with Adaptive Sampling and 3D Spherical Voxel Convolution},
  author    = {Yang You and Yujing Lou and Qi Liu and Yu-Wing Tai and Lizhuang Ma and Cewu Lu and Weiming Wang},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year      = {2020},
  eprint    = {1811.09361},
  archivePrefix = {arXiv}
}

Recommended citation: You, Y., Lou, Y., Liu, Q., Tai, Y. W., Ma, L., Lu, C., & Wang, W. (2020). Pointwise Rotation-Invariant Network with Adaptive Sampling and 3D Spherical Voxel Convolution. In AAAI (pp. 12717-12724).