Learning Hand-Eye Coordination for Robotic Grasping with Deep Learning and Large-Scale Data Collection
This addresses the problem of hand-eye coordination for robotic grasping, offering a scalable learning-based approach that is incremental over prior methods.
The paper tackles robotic grasping by training a convolutional neural network to predict grasp success from monocular images, enabling real-time control without camera calibration, and it demonstrates effective grasping of novel objects using data from over 800,000 attempts.
We describe a learning-based approach to hand-eye coordination for robotic grasping from monocular images. To learn hand-eye coordination for grasping, we trained a large convolutional neural network to predict the probability that task-space motion of the gripper will result in successful grasps, using only monocular camera images and independently of camera calibration or the current robot pose. This requires the network to observe the spatial relationship between the gripper and objects in the scene, thus learning hand-eye coordination. We then use this network to servo the gripper in real time to achieve successful grasps. To train our network, we collected over 800,000 grasp attempts over the course of two months, using between 6 and 14 robotic manipulators at any given time, with differences in camera placement and hardware. Our experimental evaluation demonstrates that our method achieves effective real-time control, can successfully grasp novel objects, and corrects mistakes by continuous servoing.