Learning a visuomotor controller for real world robotic grasping using simulated depth images

Abstract

We want to build robots that are useful in unstructured real world applications, such as doing work in the household. Grasping in particular is an important skill in this domain, yet it remains a challenge. One of the key hurdles is handling unexpected changes or motion in the objects being grasped and kinematic noise or other errors in the robot. This paper proposes an approach to learning a closed-loop controller for robotic grasping that dynamically guides the gripper to the object. We use a wrist-mounted sensor to acquire depth images in front of the gripper and train a convolutional neural network to learn a distance function to true grasps for grasp configurations over an image. The training sensor data is generated in simulation, a major advantage over previous work that uses real robot experience, which is costly to obtain. Despite being trained in simulation, our approach works well on real noisy sensor images. We compare our controller in simulated and real robot experiments to a strong baseline for grasp pose detection, and find that our approach significantly outperforms the baseline in the presence of kinematic noise, perceptual errors and disturbances of the object during grasping.

1st Conference on Robot Learning (CoRL), 13-15 November 2017, Mountain View, CA

References in corpus (2)

• Caffe: Convolutional Architecture for Fast Feature Embedding
• Active vision for dexterous grasping of novel objects

Cited by in corpus (28)

• QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulation
• How to Train Your Robot with Deep Reinforcement Learning; Lessons We've Learned
• The Feeling of Success: Does Touch Sensing Help Predict Grasp Outcomes?
• One-Shot Imitation from Observing Humans via Domain-Adaptive Meta-Learning
• Mechanical Search: Multi-Step Retrieval of a Target Object Occluded by Clutter
• Reinforcement and Imitation Learning for Diverse Visuomotor Skills
• Real-Time Deep Learning Approach to Visual Servo Control and Grasp Detection for Autonomous Robotic Manipulation
• Vision-driven Compliant Manipulation for Reliable, High-Precision Assembly Tasks
• Multi-FinGAN: Generative Coarse-To-Fine Sampling of Multi-Finger Grasps
• Multi-modal Transfer Learning for Grasping Transparent and Specular Objects
• Learning Task-Oriented Grasping for Tool Manipulation from Simulated Self-Supervision
• Sim-and-Real Reinforcement Learning for Manipulation: A Consensus-based Approach
• Multi-Task Domain Adaptation for Deep Learning of Instance Grasping from Simulation
• Adversarial Discriminative Sim-to-real Transfer of Visuo-motor Policies
• Grasping in the Wild:Learning 6DoF Closed-Loop Grasping from Low-Cost Demonstrations
• Learning Depth via Leveraging Semantics: Self-supervised Monocular Depth Estimation with Both Implicit and Explicit Semantic Guidance
• Ab Initio Particle-based Object Manipulation
• Object Pose Estimation in Robotics Revisited
• Attention-Privileged Reinforcement Learning
• VGQ-CNN: Moving Beyond Fixed Cameras and Top-Grasps for Grasp Quality Prediction
• Pixel-Attentive Policy Gradient for Multi-Fingered Grasping in Cluttered Scenes
• Sim-to-Real Deep Reinforcement Learning with Manipulators for Pick-and-place
• Emergent Hand Morphology and Control from Optimizing Robust Grasps of Diverse Objects
• Pre-training of Deep RL Agents for Improved Learning under Domain Randomization
• Comparing Task Simplifications to Learn Closed-Loop Object Picking Using Deep Reinforcement Learning
• Learning Kinematic Descriptions using SPARE: Simulated and Physical ARticulated Extendable dataset
• Accelerating Grasp Exploration by Leveraging Learned Priors
• Morphology-Agnostic Visual Robotic Control