PointNet implementation and visualization with Open3D, an open-source library that supports rapid development of software that deals with 3D data. As part of the Open3D ecosystem, this repository demonstrates how Open3D can be used for ML/DL research projects.
This repository is forked from
fxia22's PyTorch implementation.
- Added CPU support for non-cuda-enabled devices.
- Used Open3D point cloud loader for loading PointNet datasets (
- Added example for PointNet inference with Open3D Jupyter visualization
- Added example for native OpenGL visualization with Open3D (
# Install Open3D, must be v0.4.0 or above for Jupyter support pip install open3d-python # Install PyTorch # Follow: https://pytorch.org/ # Install other dependencies pip install -r requirements.txt
open3d_pointnet_inference.ipynb. All datasets and pre-trained models
shall be downloaded automatically. If you run into issues downloading files,