SpaGBOL: Spatial-Graph-Based Orientated Localisation
Tavis Shore, Oscar Mendez, Simon Hadfield

TL;DR
SpaGBOL introduces a graph-based approach with a novel dataset and GNN architecture for cross-view geo-localisation, significantly improving accuracy by leveraging spatial relationships and neighbourhood filtering.
Contribution
It presents the first graph-structured dataset, applies GNNs to exploit spatial correlations, and introduces a new retrieval filtering method based on neighbourhood bearings.
Findings
Achieves 11% top-1 accuracy improvement over previous methods.
Demonstrates 50% accuracy boost using bearing vector filtering.
Introduces a novel graph dataset for cross-view geo-localisation.
Abstract
Cross-View Geo-Localisation within urban regions is challenging in part due to the lack of geo-spatial structuring within current datasets and techniques. We propose utilising graph representations to model sequences of local observations and the connectivity of the target location. Modelling as a graph enables generating previously unseen sequences by sampling with new parameter configurations. To leverage this newly available information, we propose a GNN-based architecture, producing spatially strong embeddings and improving discriminability over isolated image embeddings. We outline SpaGBOL, introducing three novel contributions. 1) The first graph-structured dataset for Cross-View Geo-Localisation, containing multiple streetview images per node to improve generalisation. 2) Introducing GNNs to the problem, we develop the first system that exploits the correlation between node…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsGeographic Information Systems Studies · Data Management and Algorithms · Advanced Image and Video Retrieval Techniques
MethodsConvolution · Graph Neural Network · ConvNeXt · GraphSAGE
