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/

Zendy Plus

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

Zendy Tools

Zendy Open

Technologies to predict human actions are extremely important forapplications such as human robot cooperation and autonomous driving. However, amajority of the existing algorithms focus on exploiting visual features of thevideos and do not consider the mining of relationships, which include spatialrelationships between human and scene elements as well as causal relationshipsin temporal action sequences. In fact, human beings are good at using spatialand causal relational reasoning mechanism to predict the actions of others.Inspired by this idea, we proposed a Spatial and Causal Relationship basedGraph Reasoning Network (SCR-Graph), which can be used to predict human actionsby modeling the action-scene relationship, and causal relationship betweenactions, in spatial and temporal dimensions respectively. Here, in spatialdimension, a hierarchical graph attention module is designed by iterativelyaggregating the features of different kinds of scene elements in differentlevel. In temporal dimension, we designed a knowledge graph based causalreasoning module and map the past actions to temporal causal features throughDiffusion RNN. Finally, we integrated the causality features into theheterogeneous graph in the form of shadow node, and introduced a self-attentionmodule to determine the time when the knowledge graph information should beactivated. Extensive experimental results on the VIRAT datasets demonstrate thefavorable performance of the proposed framework.

SCR-Graph: Spatial-Causal Relationships based Graph Reasoning Network for Human Action Prediction