Domain Randomization with ROS and Gazebo simulation

Final result we have obtained by applying domain randomization with a Fetch Gazebo simulation.

We have replicated the whole system required to do it: from the plugin to launch object distractors, to the grasping routine, including the deep learning training

We are going to release this code as a ROSject (www.rosjects.com) ready to be executed with a notebook with explanations about how to use it.

Comment below if you want to have the ROSject