Saved in:
Bibliographic Details
Main Authors: Zhang, Shengjun, Fei, Xin, Duan, Yueqi
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2403.19220
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914731902631936
author Zhang, Shengjun
Fei, Xin
Duan, Yueqi
author_facet Zhang, Shengjun
Fei, Xin
Duan, Yueqi
contents Point clouds captured by different sensors such as RGB-D cameras and LiDAR possess non-negligible domain gaps. Most existing methods design different network architectures and train separately on point clouds from various sensors. Typically, point-based methods achieve outstanding performances on even-distributed dense point clouds from RGB-D cameras, while voxel-based methods are more efficient for large-range sparse LiDAR point clouds. In this paper, we propose geometry-to-voxel auxiliary learning to enable voxel representations to access point-level geometric information, which supports better generalisation of the voxel-based backbone with additional interpretations of multi-sensor point clouds. Specifically, we construct hierarchical geometry pools generated by a voxel-guided dynamic point network, which efficiently provide auxiliary fine-grained geometric information adapted to different stages of voxel features. We conduct experiments on joint multi-sensor datasets to demonstrate the effectiveness of GeoAuxNet. Enjoying elaborate geometric information, our method outperforms other models collectively trained on multi-sensor datasets, and achieve competitive results with the-state-of-art experts on each single dataset.
format Preprint
id arxiv_https___arxiv_org_abs_2403_19220
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle GeoAuxNet: Towards Universal 3D Representation Learning for Multi-sensor Point Clouds
Zhang, Shengjun
Fei, Xin
Duan, Yueqi
Computer Vision and Pattern Recognition
Point clouds captured by different sensors such as RGB-D cameras and LiDAR possess non-negligible domain gaps. Most existing methods design different network architectures and train separately on point clouds from various sensors. Typically, point-based methods achieve outstanding performances on even-distributed dense point clouds from RGB-D cameras, while voxel-based methods are more efficient for large-range sparse LiDAR point clouds. In this paper, we propose geometry-to-voxel auxiliary learning to enable voxel representations to access point-level geometric information, which supports better generalisation of the voxel-based backbone with additional interpretations of multi-sensor point clouds. Specifically, we construct hierarchical geometry pools generated by a voxel-guided dynamic point network, which efficiently provide auxiliary fine-grained geometric information adapted to different stages of voxel features. We conduct experiments on joint multi-sensor datasets to demonstrate the effectiveness of GeoAuxNet. Enjoying elaborate geometric information, our method outperforms other models collectively trained on multi-sensor datasets, and achieve competitive results with the-state-of-art experts on each single dataset.
title GeoAuxNet: Towards Universal 3D Representation Learning for Multi-sensor Point Clouds
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2403.19220