Tensorflow and Pytorch implementations for pose predictions from point clouds. The Tensorflow implementation is a simplified version of the one used for producing the results for "Cutting Pose Prediction from Point Clouds". The pytorch implementation is used for ongoing work on GNNs.
- edit train.py to select between datasets and network_type (optional)
- python3 train.py
- python3 test.py
- python3 evaluate.py
- write more instructions
- Add visualization scripts