/grasp_multiObject_multiGrasp

An implementation of our RA-L work 'Real-world Multi-object, Multi-grasp Detection'

Primary LanguagePython

grasp_multiObject_multiGrasp

This is the implementation of our RA-L work 'Real-world Multi-object, Multi-grasp Detection'. The detector takes RGB-D image input and predicts multiple grasp candidates for a single object or multiple objects, in a single shot. The original arxiv paper can be found here. The final version will be updated after publication process.

If you find it helpful for your research, please consider citing:

@inproceedings{chu2018deepgrasp,
Author = {Fu-Jen Chu and Ruinian Xu and Patricio A. Vela},
Title = {Deep Grasp: Detection and Localization of Grasps with Deep Neural Networks},
booktitle = {arXiv:1802.00520},
Year = {2018}
}

If you encounter any questions, please contact me at fujenchu[at]gatech[dot]edu

Demo

  1. Clone this repository
git clone https://github.com/ivalab/grasp_multiObject_multiGrasp.git
cd grasp_multiObject_multiGrasp
  1. Build Cython modules
cd lib
make clean
make
cd ..
  1. Install Python COCO API
cd data
git clone https://github.com/pdollar/coco.git
cd coco/PythonAPI
make
cd ../../..
  1. Download pretrained models
  • trained model for grasp on dropbox drive
  • put under output/res50/train/default/
  1. Run demo
./tools/demo_graspRGD.py --net res50 --dataset grasp

you can see images pop out.

Train

  1. Generate data
    1-1. Download Cornell Dataset
    1-2. Run dataPreprocessingTest_fasterrcnn_split.m (please modify paths according to your structure)
    1-3. Follow 'Format Your Dataset' section here to check if your data follows VOC format

  2. Train

./experiments/scripts/train_faster_rcnn.sh 0 graspRGB res50

Acknowledgment

This repo borrows tons of code from

Resources