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

Establishing dense correspondences between a pair of images is an importantand general problem, covering geometric matching, optical flow and semanticcorrespondences. While these applications share fundamental challenges, such aslarge displacements, pixel-accuracy, and appearance changes, they are currentlyaddressed with specialized network architectures, designed for only oneparticular task. This severely limits the generalization capabilities of suchnetworks to new scenarios, where e.g. robustness to larger displacements orhigher accuracy is required.  In this work, we propose a universal network architecture that is directlyapplicable to all the aforementioned dense correspondence problems. We achieveboth high accuracy and robustness to large displacements by investigating thecombined use of global and local correlation layers. We further propose anadaptive resolution strategy, allowing our network to operate on virtually anyinput image resolution. The proposed GLU-Net achieves state-of-the-artperformance for geometric and semantic matching as well as optical flow, whenusing the same network and weights. Code and trained models are available athttps://github.com/PruneTruong/GLU-Net.

GLU-Net: Global-Local Universal Network for Dense Flow and Correspondences