Hurry up Yuxuan!
The long-term Goal of this repo is to build a benchmark for point cloud registration and make it easy to use.
- spilit training file into train.py, train_model.py and trainer.py
- update experiment dir code
- multi-gpu training and testing
- make regformer more flexible
- Unify dataloder format
- Long time
- KITTI Odometry
PCR = Point Cloud Registration PC = Point Cloud
- BUFFER(CVPR 2023)
- RegFormer(CVPR 2023)
- SphereFormer(CVPR 2023)
- Sprin(TPAMI 2021)