3DFeat-Net: Weakly Supervised Local 3D Features for Point Cloud Registration

@article{Yew20183DFeatNetWS,
  title={3DFeat-Net: Weakly Supervised Local 3D Features for Point Cloud Registration},
  author={Zi Jian Yew and Gim Hee Lee},
  journal={ArXiv},
  year={2018},
  volume={abs/1807.09413}
}
  • Zi Jian Yew, Gim Hee Lee
  • Published 2018
  • Computer Science
  • ArXiv
  • In this paper, we propose the 3DFeat-Net which learns both 3D feature detector and descriptor for point cloud matching using weak supervision. [...] Key Result We create training and benchmark outdoor Lidar datasets, and experiments show that 3DFeat-Net obtains state-of-the-art performance on these gravity-aligned datasets.Expand Abstract
    91 Citations

    Figures, Tables, and Topics from this paper

    A learned detector method for point cloud registration
    • Highly Influenced
    USIP: Unsupervised Stable Interest Point Detection From 3D Point Clouds
    • J. Li, Gim Hee Lee
    • Computer Science
    • 2019 IEEE/CVF International Conference on Computer Vision (ICCV)
    • 2019
    • 22
    • PDF
    CorsNet: 3D Point Cloud Registration by Deep Neural Network
    • 4
    The Perfect Match: 3D Point Cloud Matching With Smoothed Densities
    • 73
    • Highly Influenced
    • PDF
    DeepCLR: Correspondence-Less Architecture for Deep End-to-End Point Cloud Registration
    • 2
    • PDF
    3DPMNet: Plane Segmentation and Matching for Point Cloud Registration
    • Hang Yang, Hui Kong
    • Computer Science
    • 2020 3rd International Conference on Unmanned Systems (ICUS)
    • 2020
    PointLoc: Deep Pose Regressor for LiDAR Point Cloud Localization
    • PDF
    Learning Multiview 3D Point Cloud Registration
    • 19
    • PDF
    SpinNet: Learning a General Surface Descriptor for 3D Point Cloud Registration
    • Highly Influenced
    • PDF

    References

    SHOWING 1-10 OF 37 REFERENCES
    COMPARISON OF 3D INTEREST POINT DETECTORS AND DESCRIPTORS FOR POINT CLOUD FUSION
    • 63
    • PDF
    3D Point Cloud Registration for Localization Using a Deep Neural Network Auto-Encoder
    • 95
    • Highly Influential
    • PDF
    3DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructions
    • 298
    • Highly Influential
    • PDF
    Aligning point cloud views using persistent feature histograms
    • 557
    • Highly Influential
    • PDF
    PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation
    • 3,515
    • PDF
    Learning a Descriptor-Specific 3D Keypoint Detector
    • 29
    • PDF
    PPFNet: Global Context Aware Local Features for Robust 3D Point Matching
    • 140
    • PDF
    Unique shape context for 3d data description
    • 211
    • Highly Influential
    • PDF
    Registration with the Point Cloud Library: A Modular Framework for Aligning in 3-D
    • 175
    • PDF
    Fast Point Feature Histograms (FPFH) for 3D registration
    • 1,981
    • Highly Influential
    • PDF