Chris Choy*, Wei Dong*, Vladlen Koltun

CVPR 2020 (Oral)

Abstract

We present Deep Global Registration, a differentiable framework for pairwise registration of real-world 3D scans.

Deep global registration is based on three modules:

  • a 6-dimensional convolutional network for correspondence confidence prediction,
  • a differentiable Weighted Procrustes algorithm for closed-form pose estimation,
  • and a robust gradient-based SE(3) optimizer for pose refinement.

Experiments demonstrate that our approach outperforms state-of-the-art methods, both learning-based and classical, on real-world data.

Citation

@inproceedings{choy2020deep,
  title={Deep Global Registration},
  author={Choy, Christopher and Dong, Wei and Koltun, Vladlen},
  booktitle={CVPR},
  year={2020}
}

Updated: