English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Image retrieval-based cross-view localization methods often lead to verycoarse camera pose estimation, due to the limited sampling density of thedatabase satellite images. In this paper, we propose a method to increase theaccuracy of a ground camera's location and orientation by estimating therelative rotation and translation between the ground-level image and itsmatched/retrieved satellite image. Our approach designs a geometry-guidedcross-view transformer that combines the benefits of conventional geometry andlearnable cross-view transformers to map the ground-view observations to anoverhead view. Given the synthesized overhead view and observed satellitefeature maps, we construct a neural pose optimizer with strong globalinformation embedding ability to estimate the relative rotation between them.After aligning their rotations, we develop an uncertainty-guided spatialcorrelation to generate a probability map of the vehicle locations, from whichthe relative translation can be determined. Experimental results demonstratethat our method significantly outperforms the state-of-the-art. Notably, thelikelihood of restricting the vehicle lateral pose to be within 1m of itsGround Truth (GT) value on the cross-view KITTI dataset has been improved from$35.54\%$ to $76.44\%$, and the likelihood of restricting the vehicleorientation to be within $1^{\circ}$ of its GT value has been improved from$19.64\%$ to $99.10\%$.

Boosting 3-DoF Ground-to-Satellite Camera Localization Accuracy via Geometry-Guided Cross-View Transformer