DRL Grasping
Deep Reinforcement Learning for robotic grasping from 3D octree observations using Ignition Gazebo and ROS 2. DRL Grasping implements Deep Reinforcement Learning for autonomous robotic grasping using 3D octree observations. The system leverages Ignition Gazebo for simulation and ROS 2 for robot control, enabling agents to learn dexterous grasping strategies from volumetric representations of the scene. Links