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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

We study an important, yet largely unexplored problem of large-scalecross-modal visual localization by matching ground RGB images to ageo-referenced aerial LIDAR 3D point cloud (rendered as depth images). Priorworks were demonstrated on small datasets and did not lend themselves toscaling up for large-scale applications. To enable large-scale evaluation, weintroduce a new dataset containing over 550K pairs (covering 143 km^2 area) ofRGB and aerial LIDAR depth images. We propose a novel joint embedding basedmethod that effectively combines the appearance and semantic cues from bothmodalities to handle drastic cross-modal variations. Experiments on theproposed dataset show that our model achieves a strong result of a median rankof 5 in matching across a large test set of 50K location pairs collected from a14km^2 area. This represents a significant advancement over prior works inperformance and scale. We conclude with qualitative results to highlight thechallenging nature of this task and the benefits of the proposed model. Ourwork provides a foundation for further research in cross-modal visuallocalization.

RGB2LIDAR: Towards Solving Large-Scale Cross-Modal Visual Localization