Unsupervised Reinforcement Learning for Transferable Manipulation Skill Discovery

04/29/2022
by   Daesol Cho, et al.
0

Current reinforcement learning (RL) in robotics often experiences difficulty in generalizing to new downstream tasks due to the innate task-specific training paradigm. To alleviate it, unsupervised RL, a framework that pre-trains the agent in a task-agnostic manner without access to the task-specific reward, leverages active exploration for distilling diverse experience into essential skills or reusable knowledge. For exploiting such benefits also in robotic manipulation, we propose an unsupervised method for transferable manipulation skill discovery that ties structured exploration toward interacting behavior and transferable skill learning. It not only enables the agent to learn interaction behavior, the key aspect of the robotic manipulation learning, without access to the environment reward, but also to generalize to arbitrary downstream manipulation tasks with the learned task-agnostic skills. Through comparative experiments, we show that our approach achieves the most diverse interacting behavior and significantly improves sample efficiency in downstream tasks including the extension to multi-object, multitask problems.

READ FULL TEXT

page 1

page 5

page 6

page 7

research
02/02/2022

Lipschitz-constrained Unsupervised Skill Discovery

We study the problem of unsupervised skill discovery, whose goal is to l...
research
08/04/2021

Learning Task Agnostic Skills with Data-driven Guidance

To increase autonomy in reinforcement learning, agents need to learn use...
research
09/24/2022

Accelerating Reinforcement Learning for Autonomous Driving using Task-Agnostic and Ego-Centric Motion Skills

Efficient and effective exploration in continuous space is a central pro...
research
10/29/2020

"What, not how": Solving an under-actuated insertion task from scratch

Robot manipulation requires a complex set of skills that need to be care...
research
09/30/2022

ASPiRe:Adaptive Skill Priors for Reinforcement Learning

We introduce ASPiRe (Adaptive Skill Prior for RL), a new approach that l...
research
04/10/2017

Stochastic Neural Networks for Hierarchical Reinforcement Learning

Deep reinforcement learning has achieved many impressive results in rece...
research
06/22/2022

Curious Exploration via Structured World Models Yields Zero-Shot Object Manipulation

It has been a long-standing dream to design artificial agents that explo...

Please sign up or login with your details

Forgot password? Click here to reset