Grasping enables the robots to manipulate obstacles or change its state on the workspace. Grasp detection is a basic skill required for robots to perform grasping and manipulation tasks in the workspace. This project demonstrates a two-staged framework where an object is detected and the grasp prediciton is done within that area. The figure below shows the yolo-ggcnn framework.
This code is tested with Python 3.10.12 on Linux Mint 21.3 and CoppeliaSim 4.6.
- Clone this repository.
- In CoppeliaSim simulator open the scene file
yolo-ggcnn.tttlocated in assets folder. - Run the
run_yolo.pyscript. - Run the
run_ggcnn.pyscript.