RoboLoc: A Benchmark Dataset for Point Place Recognition and Localization in Indoor-Outdoor Integrated Environments

TL;DR

RoboLoc is a new benchmark dataset that enables evaluation of LiDAR-based place recognition and localization across indoor-outdoor transitions, addressing the lack of existing datasets for seamless domain shift analysis.

Contribution

We introduce RoboLoc, a comprehensive dataset for GPS-free place recognition in environments with indoor-outdoor transitions, and benchmark multiple state-of-the-art models on it.

Findings

State-of-the-art models show significant domain shift challenges.

RoboLoc captures diverse elevation profiles and transition scenarios.

Benchmark results highlight the need for more robust multi-domain localization methods.

Abstract

Robust place recognition is essential for reliable localization in robotics, particularly in complex environments with frequent indoor-outdoor transitions. However, existing LiDAR-based datasets often focus on outdoor scenarios and lack seamless domain shifts. In this paper, we propose RoboLoc, a benchmark dataset designed for GPS-free place recognition in indoor-outdoor environments with floor transitions. RoboLoc features real-world robot trajectories, diverse elevation profiles, and transitions between structured indoor and unstructured outdoor domains. We benchmark a variety of state-of-the-art models, point-based, voxel-based, and BEV-based architectures, highlighting their generalizability domain shifts. RoboLoc provides a realistic testbed for developing multi-domain localization systems in robotics and autonomous navigation