Accelerating Robotic Reinforcement Learning via Parameterized Action Primitives

TL;DR

This paper introduces robot action primitives (RAPS) that are parameterized and learned by RL policies, significantly improving learning efficiency and performance across diverse robotic tasks with image inputs and sparse rewards.

Contribution

The work proposes a novel interface of RL with robots using parameterized action primitives, enhancing exploration and transferability compared to prior skill learning methods.

Findings

Parameterizing primitives improves exploration efficiency.

Significant performance gains across multiple domains.

Method outperforms prior offline skill learning approaches.

Abstract

Despite the potential of reinforcement learning (RL) for building general-purpose robotic systems, training RL agents to solve robotics tasks still remains challenging due to the difficulty of exploration in purely continuous action spaces. Addressing this problem is an active area of research with the majority of focus on improving RL methods via better optimization or more efficient exploration. An alternate but important component to consider improving is the interface of the RL algorithm with the robot. In this work, we manually specify a library of robot action primitives (RAPS), parameterized with arguments that are learned by an RL policy. These parameterized primitives are expressive, simple to implement, enable efficient exploration and can be transferred across robots, tasks and environments. We perform a thorough empirical study across challenging tasks in three distinct domains with image input and a sparse terminal reward. We find that our simple change to the action interface substantially improves both the learning efficiency and task performance irrespective of the underlying RL algorithm, significantly outperforming prior methods which learn skills from offline expert data. Code and videos at https://mihdalal.github.io/raps/