Depth Completion using Geometry-Aware Embedding

被引:21
|
作者
Du, Wenchao [1 ]
Chen, Hu [1 ]
Yang, Hongyu [1 ]
Zhang, Yi [1 ]
机构
[1] Sichuan Univ, Coll Comp Sci, Chengdu 610065, Peoples R China
关键词
NETWORK; SPARSE;
D O I
10.1109/ICRA46639.2022.9811556
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Exploiting internal spatial geometric constraints of sparse LiDARs is beneficial to depth completion, however, has been not explored well. This paper proposes an efficient method to learn geometry-aware embedding, which encodes the local and global geometric structure information from 3D points, e.g., scene layout, object's sizes and shapes, to guide dense depth estimation. Specifically, we utilize the dynamic graph representation to model generalized geometric relationship from irregular point clouds in a flexible and efficient manner. Further, we joint this embedding and corresponded RGB appearance information to infer missing depths of the scene with well structure-preserved details. The key to our method is to integrate implicit 3D geometric representation into a 2D learning architecture, which leads to a better trade-off between the performance and efficiency. Extensive experiments demonstrate that the proposed method outperforms previous works and could reconstruct fine depths with crisp boundaries in regions that are over-smoothed by them. The ablation study gives more insights into our method that could achieve significant gains with a simple design, while having better generalization capability and stability. The code is available at https://github.com/Wenchao-Du/GAENet.
引用
收藏
页码:8680 / 8686
页数:7
相关论文
共 50 条
  • [1] GraphCSPN: Geometry-Aware Depth Completion via Dynamic GCNs
    Liu, Xin
    Shao, Xiaofei
    Wang, Bo
    Li, Yali
    Wang, Shengjin
    COMPUTER VISION - ECCV 2022, PT XXXIII, 2022, 13693 : 90 - 107
  • [2] Tri-Perspective View Decomposition for Geometry-Aware Depth Completion
    Yan, Zhiqiang
    Lin, Yuankai
    Wang, Kun
    Zheng, Yupeng
    Wang, Yufei
    Zhang, Zhenyu
    Li, Jun
    Yang, Jian
    2024 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2024, 2024, : 4874 - 4884
  • [3] Geometry-Aware Face Completion and Editing
    Song, Linsen
    Cao, Jie
    Song, Lingxiao
    Hu, Yibo
    He, Ran
    THIRTY-THIRD AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTY-FIRST INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE / NINTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2019, : 2506 - 2513
  • [4] Geometry-Aware Planar Embedding of Treelike Structures
    Hu, Ping
    Boorboor, Saeed
    Marino, Joseph
    Kaufman, Arie E.
    IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS, 2023, 29 (07) : 3182 - 3194
  • [5] Collaborative Geometry-Aware Augmented Reality with Depth Sensors
    Ma, Yuhao
    Ferguson, Joshua
    Jonaitis, Kevin
    Boos, Kyle
    Patterson, Donald
    PROCEEDINGS OF THE 2014 ACM INTERNATIONAL JOINT CONFERENCE ON PERVASIVE AND UBIQUITOUS COMPUTING (UBICOMP'14 ADJUNCT), 2014, : 255 - 258
  • [6] PoinTr: Diverse Point Cloud Completion with Geometry-Aware Transformers
    Yu, Xumin
    Rao, Yongming
    Wang, Ziyi
    Liu, Zuyan
    Lu, Jiwen
    Zhou, Jie
    2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021), 2021, : 12478 - 12487
  • [7] A geometry-aware deep network for depth estimation in monocular endoscopy
    Yang, Yongming
    Shao, Shuwei
    Yang, Tao
    Wang, Peng
    Yang, Zhuo
    Wu, Chengdong
    Liu, Hao
    ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE, 2023, 122
  • [8] AdaPoinTr: Diverse Point Cloud Completion With Adaptive Geometry-Aware Transformers
    Yu, Xumin
    Rao, Yongming
    Wang, Ziyi
    Lu, Jiwen
    Zhou, Jie
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2023, 45 (12) : 14114 - 14130
  • [9] DCL: Differential Contrastive Learning for Geometry-Aware Depth Synthesis
    Shen, Yuefan
    Yang, Yanchao
    Zheng, Youyi
    Liu, C. Karen
    Guibas, Leonidas J.
    IEEE ROBOTICS AND AUTOMATION LETTERS, 2022, 7 (02) : 4845 - 4852
  • [10] Geometry-Aware Symmetric Domain Adaptation for Monocular Depth Estimation
    Zhao, Shanshan
    Fu, Huan
    Gong, Mingming
    Tao, Dacheng
    2019 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2019), 2019, : 9780 - 9790