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

Rigid registration of point clouds is a fundamental problem in computervision with many applications from 3D scene reconstruction to geometry captureand robotics. If a suitable initial registration is available, conventionalmethods like ICP and its many variants can provide adequate solutions. Inabsence of a suitable initialization and in the presence of a high outlier rateor in the case of small overlap though the task of rigid registration stillpresents great challenges. The advent of deep learning in computer vision hasbrought new drive to research on this topic, since it provides the possibilityto learn expressive feature-representations and provide one-shot estimatesinstead of depending on time-consuming iterations of conventional robustmethods. Yet, the rotation and permutation invariant nature of point cloudsposes its own challenges to deep learning, resulting in loss of performance andlow generalization capability due to sensitivity to outliers andcharacteristics of 3D scans not present during network training. In this work,we present a novel fast and light-weight network architecture using theattention mechanism to augment point descriptors at inference time to optimallysuit the registration task of the specific point clouds it is presented with.Employing a fully-connected graph both within and between point clouds lets thenetwork reason about the importance and reliability of points for registration,making our approach robust to outliers, low overlap and unseen data. We testthe performance of our registration algorithm on different registration andgeneralization tasks and provide information on runtime and resourceconsumption. The code and trained weights are available athttps://github.com/mordecaimalignatius/GAFAR/.

GAFAR: Graph-Attention Feature-Augmentation for Registration A Fast and Light-weight Point Set Registration Algorithm