Visit the GraspNet Website to get the dataset.

You can install using pip.

pip install graspnetAPI

You can also install from source.

git clone https://github.com/graspnet/graspnetAPI.git
cd graspnetAPI
pip install .

Refer to online document for more details.
PDF Document is available, too.

You can also build the doc manually.

cd docs
pip install -r requirements.txt
bash build_doc.sh

LaTeX is required to build the pdf, but html can be built anyway.

The frame of our gripper is defined as

cd examples

# change the path of graspnet root

# How to load labels from graspnet.
python3 exam_loadGrasp.py

# How to convert between 6d and rectangle grasps.
python3 exam_convert.py

# Check the completeness of the data.
python3 exam_check_data.py

# you can also run other examples

Please refer to our document for more examples.

Please cite these papers in your publications if it helps your research:

@inproceedings{fang2020graspnet,
  title={GraspNet-1Billion: A Large-Scale Benchmark for General Object Grasping},
  author={Fang, Hao-Shu and Wang, Chenxi and Gou, Minghao and Lu, Cewu},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR)},
  pages={11444--11453},
  year={2020}
}

• Add transformation for Grasp and GraspGroup.

• Add inpainting for depth image.

• Minor fix bug on loadScenePointCloud.