MARVEL: Multi-Agent Reinforcement Learning for constrained field-of-View multi-robot Exploration in Large-scale environments

This repository hosts the code for MARVEL, accepted for ICRA 2025.

Supplementary video link: YouTube

MARVEL is a neural framework that leverages graph attention networks, together with novel frontiers and orientation features fusion technique, to develop a collaborative, decentralized policy using multi-agent reinforcement learning (MARL) for robots with constrained FoV.

Use conda to setup your environment:

conda env create -f marvel.yml
conda activate marvel

First, set the appropriate parameters in test_parameter.py and adjust testing configurations within test_driver.py. Run test_driver.py to evaluate.

Set appropriate parameters in parameter.py and run driver.py to train the model.

If you find our work useful, please consider citing our paper:

@article{chiun2025marvel,
  title={MARVEL: Multi-Agent Reinforcement Learning for constrained field-of-View multi-robot Exploration in Large-scale environments},
  author={Chiun, Jimmy and Zhang, Shizhe and Wang, Yizhuo and Cao, Yuhong and Sartoretti, Guillaume},
  journal={arXiv preprint arXiv:2502.20217},
  year={2025}
}

Authors: Jimmy Chiun, Shizhe Zhang, Yizhuo Wang, Yuhong Cao, Guillaume Sartoretti