errrr0501 / ros2_gqcnn

Python3 module for GQ-CNN training and deployment with ROS2 integration.

Note: Python 2.x support has officially been dropped.

Berkeley AUTOLAB's GQCNN Package

Package Overview

The gqcnn Python package is for training and analysis of Grasp Quality Convolutional Neural Networks (GQ-CNNs). It is part of the ongoing Dexterity-Network (Dex-Net) project created and maintained by the AUTOLAB at UC Berkeley.

Installation and Usage

Please see the docs for installation and usage instructions.

Installation and Usage with docker

1.clone docker to your <workspace_ws>.

git clone https://github.com/errrr0501/docker_20.04_CUDA12_tf1.15

2.build it and run it.

3.clone and build realsense_ros2_wrapper.

#open a new terminal
mkdir <your realsense workspace>
cd <your realsense workspace>
git clone https://github.com/IntelRealSense/realsense-ros.git -b ros2-development
colcon build

4.make a workspace and clone gqcnn and autolab_core.

#open a new terminal
mkdir <your gqcnn workspace>
cd <your gqcnn workspace>
git clone https://github.com/errrr0501/ROS2_gqcnn.git
git clone https://github.com/errrr0501/ROS2_autolab_core.git
colcon build

5.use with your camera topic.

#open a new terminal
cd <your realsense workspace>
source install/setup.bash
ros2 launch realsense2_camera rs_launch.py depth_module.profile:=640x480x30 rgb_camera.profile:=640x480x30 align_depth.enable:=true
#open a new terminal
cd <your gqcnn workspace>
source install/setup.bash
ros2 launch gqcnn grasp_planning_service.launch.py
#open a new terminal
source install/setup.bash
python3 src/ROS2_gqcnn/gqcnn/examples/policy_camera_ros2.py

Citation

If you use any part of this code in a publication, please cite the appropriate Dex-Net publication.