PyTorch reimplementation for "KPConv: Flexible and Deformable Convolution for Point Clouds" https://arxiv.org/abs/1904.08889
This repo is implementation for KPConv(https://arxiv.org/abs/1904.08889) in pytorch.
There are still some works to be done:
collate_fn
where the neighbor indices and pooling indices are calculated, is too slow. In the tf version, the author implement 2 tensroflow C++ wrapper which is quite efficient. I am planing to write C++ extention using pytorch.
conda env create -f environment.yml
pytorch_ops
dictionary and run python setup.py install
to build and install the C++ extension for batch_find_neighbors
function.Due to the time limitation, I have just implemented the experiments on ShapeNet(classification and part segmentation) and ModelNet40.
python training_ModelNet.py[training_ShapeNetCls.py]
python training_ShapeNetPart.py
Thank @HuguesTHOMAS for sharing the tensorflow version and valuable explainations.