/PointNetLK

Primary LanguagePythonMIT LicenseMIT

PointNetLK: Point Cloud Registration using PointNet

Source Code Author: Yasuhiro Aoki

Requires:

  • PyTorch 0.4.0 (perhaps, 0.4.1 (the latest) will be OK.) and torchvision
  • NumPy
  • SciPy
  • MatPlotLib
  • ModelNet40

Main files for experiments:

  • train_classifier.py: train PointNet classifier (used for transfer learning)
  • train_pointlk.py: train PointNet-LK
  • generate_rotation.py: generate 6-dim perturbations (rotation and translation) (for testing)
  • test_pointlk.py: test PointNet-LK
  • test_icp.py: test ICP
  • result_stat.py: compute mean errors of above tests

Examples (Bash shell scripts):

  • ex1_train.sh: train PointNet classifier and transfer to PointNet-LK.
  • ex1_genrot.sh: generate perturbations for testing
  • ex1_test_pointlk.sh: test PointNet-LK
  • ex1_test_icp.sh: test ICP
  • ex1_result_stat.sh: compute mean errors of above tests

Citation

@InProceedings{yaoki2019pointnetlk,
       author = {Aoki, Yasuhiro and Goforth, Hunter and Arun Srivatsan, Rangaprasad and Lucey, Simon},
       title = {PointNetLK: Robust & Efficient Point Cloud Registration Using PointNet},
       booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
       month = {June},
       year = {2019}
}