REGRACE: A Robust and Efficient Graph-based Re-localization Algorithm using Consistency Evaluation
This addresses the challenge of efficient and robust re-localization for autonomous navigation systems, representing an incremental improvement over existing methods.
The paper tackles the problem of scalable and viewpoint-robust loop closure in large-scale navigation by introducing REGRACE, which uses LiDAR-based submaps with rotation-invariant features and graph neural networks, achieving similar accuracy to state-of-the-art methods while being twice as fast.
Loop closures are essential for correcting odometry drift and creating consistent maps, especially in the context of large-scale navigation. Current methods using dense point clouds for accurate place recognition do not scale well due to computationally expensive scan-to-scan comparisons. Alternative object-centric approaches are more efficient but often struggle with sensitivity to viewpoint variation. In this work, we introduce REGRACE, a novel approach that addresses these challenges of scalability and perspective difference in re-localization by using LiDAR-based submaps. We introduce rotation-invariant features for each labeled object and enhance them with neighborhood context through a graph neural network. To identify potential revisits, we employ a scalable bag-of-words approach, pooling one learned global feature per submap. Additionally, we define a revisit with geometrical consistency cues rather than embedding distance, allowing us to recognize far-away loop closures. Our evaluations demonstrate that REGRACE achieves similar results compared to state-of-the-art place recognition and registration baselines while being twice as fast. Code and models are publicly available.