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

In order to achieve better performance for point cloud analysis, manyresearchers apply deeper neural networks using stacked Multi-Layer-Perceptron(MLP) convolutions over irregular point cloud. However, applying dense MLPconvolutions over large amount of points (e.g. autonomous driving application)leads to inefficiency in memory and computation. To achieve high performancebut less complexity, we propose a deep-wide neural network, calledShufflePointNet, to exploit fine-grained local features and reduce redundancyin parallel using group convolution and channel shuffle operation. Unlikeconventional operation that directly applies MLPs on high-dimensional featuresof point cloud, our model goes wider by splitting features into groups inadvance, and each group with certain smaller depth is only responsible forrespective MLP operation, which can reduce complexity and allows to encode moreuseful information. Meanwhile, we connect communication between groups byshuffling groups in feature channel to capture fine-grained features. We claimthat, multi-branch method for wider neural networks is also beneficial tofeature extraction for point cloud. We present extensive experiments for shapeclassification task on ModelNet40 dataset and semantic segmentation task onlarge scale datasets ShapeNet part, S3DIS and KITTI. We further performablation study and compare our model to other state-of-the-art algorithms interms of complexity and accuracy.

Go Wider: An Efficient Neural Network for Point Cloud Analysis via Group Convolutions