Learning Multiview 3D Point Cloud Registration

@article{Gojcic2020LearningM3,
  title={Learning Multiview 3D Point Cloud Registration},
  author={Zan Gojcic and Caifa Zhou and J. Wegner and L. Guibas and Tolga Birdal},
  journal={2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2020},
  pages={1756-1766}
}
  • Zan Gojcic, Caifa Zhou, +2 authors Tolga Birdal
  • Published 2020
  • Computer Science
  • 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
  • We present a novel, end-to-end learnable, multiview 3D point cloud registration algorithm. Registration of multiple scans typically follows a two-stage pipeline: the initial pairwise alignment and the globally consistent refinement. The former is often ambiguous due to the low overlap of neighboring point clouds, symmetries and repetitive scene parts. Therefore, the latter global refinement aims at establishing the cyclic consistency across multiple scans and helps in resolving the ambiguous… CONTINUE READING

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 77 REFERENCES
    3D Local Features for Direct Pairwise Registration
    19
    Deep Closest Point: Learning Representations for Point Cloud Registration
    53
    DeepMapping: Unsupervised Map Estimation From Multiple Point Clouds
    15
    Super 4PCS Fast Global Pointcloud Registration via Smart Indexing
    222
    The Perfect Match: 3D Point Cloud Matching With Smoothed Densities
    42
    3DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructions
    244
    DeepVCP: An End-to-End Deep Neural Network for Point Cloud Registration
    33
    Learning Two-View Correspondences and Geometry Using Order-Aware Network
    20
    3DFeat-Net: Weakly Supervised Local 3D Features for Point Cloud Registration
    61