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

Direct image-to-image alignment that relies on the optimization ofphotometric error metrics suffers from limited convergence range andsensitivity to lighting conditions. Deep learning approaches has been appliedto address this problem by learning better feature representations usingconvolutional neural networks, yet still require a good initialization. In thispaper, we demonstrate that the inaccurate numerical Jacobian limits theconvergence range which could be improved greatly using learned approaches.Based on this observation, we propose a novel end-to-end network, RegNet, tolearn the optimization of image-to-image pose registration. By jointly learningfeature representation for each pixel and partial derivatives that replacehandcrafted ones (e.g., numerical differentiation) in the optimization step,the neural network facilitates end-to-end optimization. The energy landscape isconstrained on both the feature representation and the learned Jacobian, henceproviding more flexibility for the optimization as a consequence leads to morerobust and faster convergence. In a series of experiments, including a broadablation study, we demonstrate that RegNet is able to converge forlarge-baseline image pairs with fewer iterations.

RegNet: Learning the Optimization of Direct Image-to-Image Pose Registration