Learning Deep Parameterized Skills from Demonstration for Re-targetable Visuomotor Control

Abstract

Robots need to learn skills that can not only generalize across similar problems but also be directed to a specific goal. Previous methods either train a new skill for every different goal or do not infer the specific target in the presence of multiple goals from visual data. We introduce an end-to-end method that represents targetable visuomotor skills as a goal-parameterized neural network policy. By training on an informative subset of available goals with the associated target parameters, we are able to learn a policy that can zero-shot generalize to previously unseen goals. We evaluate our method in a representative 2D simulation of a button-grid and on both button-pressing and peg-insertion tasks on two different physical arms. We demonstrate that our model trained on 33% of the possible goals is able to generalize to more than 90% of the targets in the scene for both simulation and robot experiments. We also successfully learn a mapping from target pixel coordinates to a robot policy to complete a specified goal.

Preprint

References in corpus (7)

• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
• Generative Adversarial Imitation Learning
• One-Shot Visual Imitation Learning via Meta-Learning
• One-Shot Imitation Learning
• Automatic Goal Generation for Reinforcement Learning Agents
• Stochastic Gradient Methods with Layer-wise Adaptive Moments for Training of Deep Networks
• Goal-conditioned Imitation Learning

Cited by in corpus (4)

• Language-Conditioned Imitation Learning for Robot Manipulation Tasks
• Learning to Generalize Across Long-Horizon Tasks from Human Demonstrations
• Deep Reactive Planning in Dynamic Environments
• Imitation Learning of Robot Policies by Combining Language, Vision and Demonstration