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

Graph convolutional networks (GCNs) have been widely used and achievedremarkable results in skeleton-based action recognition. In GCNs, graphtopology dominates feature aggregation and therefore is the key to extractingrepresentative features. In this work, we propose a novel Channel-wise TopologyRefinement Graph Convolution (CTR-GC) to dynamically learn different topologiesand effectively aggregate joint features in different channels forskeleton-based action recognition. The proposed CTR-GC models channel-wisetopologies through learning a shared topology as a generic prior for allchannels and refining it with channel-specific correlations for each channel.Our refinement method introduces few extra parameters and significantly reducesthe difficulty of modeling channel-wise topologies. Furthermore, viareformulating graph convolutions into a unified form, we find that CTR-GCrelaxes strict constraints of graph convolutions, leading to strongerrepresentation capability. Combining CTR-GC with temporal modeling modules, wedevelop a powerful graph convolutional network named CTR-GCN which notablyoutperforms state-of-the-art methods on the NTU RGB+D, NTU RGB+D 120, andNW-UCLA datasets.

Channel-wise Topology Refinement Graph Convolution for Skeleton-Based Action Recognition