Created by Charles R. Qi, Li (Eric) Yi, Hao Su, Leonidas J. Guibas from Stanford University.
This is work is based on our paper linked here. The code release is still in an ongoing process... Stay tuned!
Current release includes TF operators (CPU and GPU), some core pointnet++ layers and a few example network models.
The TF operators are included under tf_ops, you need to compile them (check tf_xxx_compile.sh under each ops subfolder) first. Update nvcc and python path if necessary. The code is tested under TF1.2.0. If you are using earlier version it's possible that you need to remove the -D_GLIBCXX_USE_CXX11_ABI=0 flag in g++ command in order to compile correctly.
TF and pointnet++ utility layers are defined under utils/tf_util.py and utils/pointnet_util.py
Under models, two classification models (SSG and MSG) and SSG models for part and semantic segmentation have been included.
You can get our sampled point clouds of ModelNet40 (XYZ and normal from mesh, 10k points per shape) at this OneDrive link.