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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Recent advances in deep learning for 3D point clouds have shown greatpromises in scene understanding tasks thanks to the introduction of convolutionoperators to consume 3D point clouds directly in a neural network. Point clouddata, however, could have arbitrary rotations, especially those acquired from3D scanning. Recent works show that it is possible to design point cloudconvolutions with rotation invariance property, but such methods generally donot perform as well as translation-invariant only convolution. We found that akey reason is that compared to point coordinates, rotation-invariant featuresconsumed by point cloud convolution are not as distinctive. To address thisproblem, we propose a novel convolution operator that enhances featuredistinction by integrating global context information from the input pointcloud to the convolution. To this end, a globally weighted local referenceframe is constructed in each point neighborhood in which the local point set isdecomposed into bins. Anchor points are generated in each bin to representglobal shape features. A convolution can then be performed to transform thepoints and anchor features into final rotation-invariant features. We conductseveral experiments on point cloud classification, part segmentation, shaperetrieval, and normals estimation to evaluate our convolution, which achievesstate-of-the-art accuracy under challenging rotations.

Global Context Aware Convolutions for 3D Point Cloud Understanding