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

This paper studies image-based geo-localization (IBL) problem usingground-to-aerial cross-view matching. The goal is to predict the spatiallocation of a ground-level query image by matching it to a large geotaggedaerial image database (e.g., satellite imagery). This is a challenging task dueto the drastic differences in their viewpoints and visual appearances. Existingdeep learning methods for this problem have been focused on maximizing featuresimilarity between spatially close-by image pairs, while minimizing otherimages pairs which are far apart. They do so by deep feature embedding based onvisual appearance in those ground-and-aerial images. However, in everyday life,humans commonly use {\em orientation} information as an important cue for thetask of spatial localization. Inspired by this insight, this paper proposes anovel method which endows deep neural networks with the `commonsense' oforientation. Given a ground-level spherical panoramic image as query input (anda large georeferenced satellite image database), we design a Siamese networkwhich explicitly encodes the orientation (i.e., spherical directions) of eachpixel of the images. Our method significantly boosts the discriminative powerof the learned deep features, leading to a much higher recall and precisionoutperforming all previous methods. Our network is also more compact using only1/5th number of parameters than a previously best-performing network. Toevaluate the generalization of our method, we also created a large-scalecross-view localization benchmark containing 100K geotagged ground-aerial pairscovering a city. Our codes and datasets are available at\url{https://github.com/Liumouliu/OriCNN}.

Lending Orientation to Neural Networks for Cross-view Geo-localization