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

The analyses relying on 3D point clouds are an utterly complex task, ofteninvolving million of points, but also requiring computationally efficientalgorithms because of many real-time applications; e.g. autonomous vehicle.However, point clouds are intrinsically irregular and the points are sparselydistributed in a non-Euclidean space, which normally requires point-wiseprocessing to achieve high performances. Although shared filter matrices andpooling layers in convolutional neural networks (CNNs) are capable of reducingthe dimensionality of the problem and extracting high-level informationsimultaneously, grids and highly regular data format are required as input. Inorder to balance model performance and complexity, we introduce a novel neuralnetwork architecture exploiting local features from a manually subsampled pointset. In our network, a recursive farthest point sampling method is firstlyapplied to efficiently cover the entire point set. Successively, we employ thek-nearest neighbours (knn) algorithm to gather local neighbourhood for eachgroup of the subsampled points. Finally, a multiple layer perceptron (MLP) isapplied on the subsampled points and edges that connect corresponding point andneighbours to extract local features. The architecture has been tested for bothshape classification and segmentation using the ModelNet40 and ShapeNet partdatasets, in order to show that the network achieves the best trade-off interms of competitive performance when compared to other state-of-the-artalgorithms.

Fast Hierarchical Neural Network for Feature Learning on Point Cloud