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/

Zendy Plus

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

Zendy Tools

Zendy Open

Low-overlap regions between paired point clouds make the captured featuresvery low-confidence, leading cutting edge models to point cloud registrationwith poor quality. Beyond the traditional wisdom, we raise an intriguingquestion: Is it possible to exploit an intermediate yet misaligned imagebetween two low-overlap point clouds to enhance the performance of cutting-edgeregistration models? To answer it, we propose a misaligned image supportedregistration network for low-overlap point cloud pairs, dubbed ImLoveNet.ImLoveNet first learns triple deep features across different modalities andthen exports these features to a two-stage classifier, for progressivelyobtaining the high-confidence overlap region between the two point clouds.Therefore, soft correspondences are well established on the predicted overlapregion, resulting in accurate rigid transformations for registration. ImLoveNetis simple to implement yet effective, since 1) the misaligned image providesclearer overlap information for the two low-overlap point clouds to betterlocate overlap parts; 2) it contains certain geometry knowledge to extractbetter deep features; and 3) it does not require the extrinsic parameters ofthe imaging device with respect to the reference frame of the 3D point cloud.Extensive qualitative and quantitative evaluations on different kinds ofbenchmarks demonstrate the effectiveness and superiority of our ImLoveNet overstate-of-the-art approaches.

ImLoveNet: Misaligned Image-supported Registration Network for Low-overlap Point Cloud Pairs