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

Unlike on images, semantic learning on 3D point clouds using a deep networkis challenging due to the naturally unordered data structure. Among existingworks, PointNet has achieved promising results by directly learning on pointsets. However, it does not take full advantage of a point's local neighborhoodthat contains fine-grained structural information which turns out to be helpfultowards better semantic learning. In this regard, we present two new operationsto improve PointNet with a more efficient exploitation of local structures. Thefirst one focuses on local 3D geometric structures. In analogy to a convolutionkernel for images, we define a point-set kernel as a set of learnable 3D pointsthat jointly respond to a set of neighboring data points according to theirgeometric affinities measured by kernel correlation, adapted from a similartechnique for point cloud registration. The second one exploits localhigh-dimensional feature structures by recursive feature aggregation on anearest-neighbor-graph computed from 3D positions. Experiments show that ournetwork can efficiently capture local information and robustly achieve betterperformances on major datasets. Our code is available athttp://www.merl.com/research/license#KCNet

Mining Point Cloud Local Structures by Kernel Correlation and Graph Pooling