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

Deep learning with 3D data has progressed significantly since theintroduction of convolutional neural networks that can handle point orderambiguity in point cloud data. While being able to achieve good accuracies invarious scene understanding tasks, previous methods often have low trainingspeed and complex network architecture. In this paper, we address theseproblems by proposing an efficient end-to-end permutation invariant convolutionfor point cloud deep learning. Our simple yet effective convolution operatornamed ShellConv uses statistics from concentric spherical shells to definerepresentative features and resolve the point order ambiguity, allowingtraditional convolution to perform on such features. Based on ShellConv wefurther build an efficient neural network named ShellNet to directly consumethe point clouds with larger receptive fields while maintaining less layers. Wedemonstrate the efficacy of ShellNet by producing state-of-the-art results onobject classification, object part segmentation, and semantic scenesegmentation while keeping the network very fast to train.

ShellNet: Efficient Point Cloud Convolutional Neural Networks using Concentric Shells Statistics