Active Perception for Grasp Detection via Neural Graspness Field
Paper

Haoxiang Ma, Modi Shi, Boyang Gao, Di Huang
In NeurIPS'2024

This repository is still under development and I will update it gradually in few weeks.

This repository is official PyTorch implementation for our NeurIPS2024 paper. The code is based on ESLAM

Get the code.

git clone https://github.com/mahaoxiang822/ActiveNGF.git
cd ActiveNGF

Install packages via Pip.

pip install -r requirements.txt

Compile and install pointnet2 operators (code adapted from votenet).

cd pointnet2
python setup.py install

Download the pretrained graspness prediction model from Google Drive and put it under

ckpts/
---|graspness.tar

Using the scene config file in config/ to run the pipeline. As an example, to run ActiveNGF on for the scene_0100, run:

python -W ignore run.py configs/GraspNet/scene_0100.yaml

update evaluation code

You can contact the author through email: mahaoxiang822@buaa.edu.cn

If you find our work useful, please consider citing:

@inproceedings{ma2024activengf,
  author       = {Haoxiang Ma and
                  Modi Shi and
                  Boyang Gao and
                  Di Huang},
  title        = {Active Perception for Grasp Detection via Neural Graspness Field},
  booktitle    = {Annual Conference on Neural Information Processing Systems},
  year         = {2024},
}