In this paper, we propose 3D point-capsule networks, an auto-encoder designed to process sparse 3D point clouds while preserving spatial arrangements of the input data. 3D capsule networks arise as a direct consequence of our novel unified 3D auto-encoder formulation. Their dynamic routing scheme and the peculiar 2D latent space deployed by our approach bring in improvements for several common point cloud-related tasks, such as object classification, object reconstruction and part segmentation as substantiated by our extensive evaluations. Moreover, it enables new applications such as part interpolation and replacement.
@article{arxiv.1812.10775,
title = {3D Point Capsule Networks},
author = {Yongheng Zhao and Tolga Birdal and Haowen Deng and Federico Tombari},
journal= {arXiv preprint arXiv:1812.10775},
year = {2019}
}
Comments
As published in CVPR 2019 (camera ready version), with supplementary material