ESP: Exploiting Symmetry Prior for Multi-Agent Reinforcement Learning

TL;DR

This paper introduces a symmetry-based framework for multi-agent reinforcement learning that enhances data efficiency and generalization by leveraging prior knowledge through data augmentation and consistency loss.

Contribution

It proposes a model-agnostic framework exploiting symmetry priors to improve data efficiency in MARL, applicable across various algorithms.

Findings

Effective in multiple challenging tasks

Improves data efficiency and generalization

Proven superior in multi-robot experiments

Abstract

Multi-agent reinforcement learning (MARL) has achieved promising results in recent years. However, most existing reinforcement learning methods require a large amount of data for model training. In addition, data-efficient reinforcement learning requires the construction of strong inductive biases, which are ignored in the current MARL approaches. Inspired by the symmetry phenomenon in multi-agent systems, this paper proposes a framework for exploiting prior knowledge by integrating data augmentation and a well-designed consistency loss into the existing MARL methods. In addition, the proposed framework is model-agnostic and can be applied to most of the current MARL algorithms. Experimental tests on multiple challenging tasks demonstrate the effectiveness of the proposed framework. Moreover, the proposed framework is applied to a physical multi-robot testbed to show its superiority.