G2IA: Geometry-Guided Instance-Aware Retrieval and Refinement for Cross-Modal Place Recognition

2026-06-13

AI Summary

This work addresses the modality gap between images and point clouds and the ambiguity in urban scene perception for cross-modal place recognition. The authors propose a two-stage framework: first, they integrate visual geometric priors with instance-level features to construct a global descriptor compatible with LiDAR maps for efficient retrieval; then, they refine the results via fine-grained re-ranking based on cross-modal consistency in instance shape and spatial layout. By introducing a geometry-guided, instance-aware mechanism, the method effectively combines global retrieval with local verification, significantly enhancing the reliability and generalization of cross-modal alignment. Evaluated on multiple public benchmarks, the approach achieves state-of-the-art accuracy in image-to-point-cloud place recognition and demonstrates strong cross-dataset generalization performance.

Abstract

Cross-modal place recognition (CMPR) enables camera-only robots to localize against pre-built LiDAR maps in autonomous navigation scenarios. This image-to-point-cloud setting is challenged by two coupled ambiguities: the modality gap between perspective RGB appearance and sparse metric geometry, and perceptual aliasing among urban places with similar roads, facades, intersections, and object arrangements. Instead of treating CMPR as a single global descriptor matching problem, we argue that reliable retrieval requires both geometry-aware representation alignment and fine-grained candidate verification. In this paper, we propose G2IA, a geometry-guided instance-aware framework for image-to-point-cloud place recognition. In the retrieval stage, visual geometry priors from VGGT and instance features are integrated to construct place descriptors that are more compatible with LiDAR-derived map representations. In the refinement stage, the retrieved candidates are re-ranked by explicitly verifying whether local instance shapes and their relative spatial layouts are consistent across modalities. Experiments on public benchmarks demonstrate that G2IA consistently improves image-to-point-cloud place recognition under different localization thresholds, and exhibits strong cross-dataset generalization.

Problem

• cross-modal place recognition
• modality gap
• perceptual aliasing
• image-to-point-cloud
• autonomous navigation

Innovation

• geometry-guided
• instance-aware
• cross-modal place recognition
• descriptor alignment
• spatial layout verification