PEnG: Pose-Enhanced Geo-Localisation

TL;DR

PEnG introduces a two-stage system combining graph-based localization and relative pose estimation to significantly improve cross-view geo-localisation accuracy, achieving sub-metre precision and outperforming previous methods.

Contribution

The paper presents PEnG, the first technique to utilize both viewpoints in cross-view geo-localisation datasets for centimeter-level accuracy.

Findings

Median error reduced by 96.9% to 22.77m.

Relative Top-5m retrieval improved by 213%.

Achieved sub-metre to centimetre-level localization accuracy.

Abstract

Cross-view Geo-localisation is typically performed at a coarse granularity, because densely sampled satellite image patches overlap heavily. This heavy overlap would make disambiguating patches very challenging. However, by opting for sparsely sampled patches, prior work has placed an artificial upper bound on the localisation accuracy that is possible. Even a perfect oracle system cannot achieve accuracy greater than the average separation of the tiles. To solve this limitation, we propose combining cross-view geo-localisation and relative pose estimation to increase precision to a level practical for real-world application. We develop PEnG, a 2-stage system which first predicts the most likely edges from a city-scale graph representation upon which a query image lies. It then performs relative pose estimation within these edges to determine a precise position. PEnG presents the first technique to utilise both viewpoints available within cross-view geo-localisation datasets to enhance precision to a sub-metre level, with some examples achieving centimetre level accuracy. Our proposed ensemble achieves state-of-the-art precision - with relative Top-5m retrieval improvements on previous works of 213%. Decreasing the median euclidean distance error by 96.90% from the previous best of 734m down to 22.77m, when evaluating with 90 degree horizontal FOV images. Code will be made available: tavisshore.co.uk/PEnG