Robustness via Retrying: Closed-Loop Robotic Manipulation with Self-Supervised Learning

Abstract

Prediction is an appealing objective for self-supervised learning of behavioral skills, particularly for autonomous robots. However, effectively utilizing predictive models for control, especially with raw image inputs, poses a number of major challenges. How should the predictions be used? What happens when they are inaccurate? In this paper, we tackle these questions by proposing a method for learning robotic skills from raw image observations, using only autonomously collected experience. We show that even an imperfect model can complete complex tasks if it can continuously retry, but this requires the model to not lose track of the objective (e.g., the object of interest). To enable a robot to continuously retry a task, we devise a self-supervised algorithm for learning image registration, which can keep track of objects of interest for the duration of the trial. We demonstrate that this idea can be combined with a video-prediction based controller to enable complex behaviors to be learned from scratch using only raw visual inputs, including grasping, repositioning objects, and non-prehensile manipulation. Our real-world experiments demonstrate that a model trained with 160 robot hours of autonomously collected, unlabeled data is able to successfully perform complex manipulation tasks with a wide range of objects not seen during training.

arXiv commentaccepted at the Conference on Robot Learning (CoRL) 2018

Cited by in corpus (17)

• Revisiting Self-Supervised Visual Representation Learning
• Entity Abstraction in Visual Model-Based Reinforcement Learning
• FitVid: Overfitting in Pixel-Level Video Prediction
• Combining Coarse and Fine Physics for Manipulation using Parallel-in-Time Integration
• Model-Based Visual Planning with Self-Supervised Functional Distances
• Contextual Imagined Goals for Self-Supervised Robotic Learning
• Physion: Evaluating Physical Prediction from Vision in Humans and Machines
• Self-Supervised Learning of State Estimation for Manipulating Deformable Linear Objects
• Example-Driven Model-Based Reinforcement Learning for Solving Long-Horizon Visuomotor Tasks
• Greedy Hierarchical Variational Autoencoders for Large-Scale Video Prediction
• Towards a Hypothesis on Visual Transformation based Self-Supervision
• Distribution Estimation to Automate Transformation Policies for Self-Supervision
• Learning to Manipulate Object Collections Using Grounded State Representations
• Distributed Reinforcement Learning of Targeted Grasping with Active Vision for Mobile Manipulators
• Robot Perception enables Complex Navigation Behavior via Self-Supervised Learning
• Learning Topological Motion Primitives for Knot Planning
• Depth by Poking: Learning to Estimate Depth from Self-Supervised Grasping