RHAML: Rendezvous-based Hierarchical Architecture for Mutual Localization

TL;DR

RHAML introduces a hierarchical architecture with multi-scale feature learning and pose refinement to improve mutual localization accuracy in multi-robot systems, especially under large scale variations, achieving high precision in simulations and practical mapping tasks.

Contribution

The paper presents a novel rendezvous-based hierarchical architecture that enhances mutual localization accuracy using anisotropic convolutions, iterative pose refinement, and global pose graph optimization.

Findings

Achieves translation errors below 2 cm in simulations.

Achieves rotation errors below 0.5 degrees in simulations.

Validates utility in multi-robot map fusion in unknown environments.

Abstract

Mutual localization serves as the foundation for collaborative perception and task assignment in multi-robot systems. Effectively utilizing limited onboard sensors for mutual localization between marker-less robots is a worthwhile goal. However, due to inadequate consideration of large scale variations of the observed robot and localization refinement, previous work has shown limited accuracy when robots are equipped only with RGB cameras. To enhance the precision of localization, this paper proposes a novel rendezvous-based hierarchical architecture for mutual localization (RHAML). Firstly, to learn multi-scale robot features, anisotropic convolutions are introduced into the network, yielding initial localization results. Then, the iterative refinement module with rendering is employed to adjust the observed robot poses. Finally, the pose graph is conducted to globally optimize all localization results, which takes into account multi-frame observations. Therefore, a flexible architecture is provided that allows for the selection of appropriate modules based on requirements. Simulations demonstrate that RHAML effectively addresses the problem of multi-robot mutual localization, achieving translation errors below 2 cm and rotation errors below 0.5 degrees when robots exhibit 5 m of depth variation. Moreover, its practical utility is validated by applying it to map fusion when multi-robots explore unknown environments.