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

Analyzing and forecasting trajectories of agents like pedestrians and cars incomplex scenes has become more and more significant in many intelligent systemsand applications. The diversity and uncertainty in socially interactivebehaviors among a rich variety of agents make this task more challenging thanother deterministic computer vision tasks. Researchers have made a lot ofefforts to quantify the effects of these interactions on future trajectoriesthrough different mathematical models and network structures, but this problemhas not been well solved. Inspired by marine animals that localize thepositions of their companions underwater through echoes, we build a newanglebased trainable social representation, named SocialCircle, forcontinuously reflecting the context of social interactions at different angularorientations relative to the target agent. We validate the effect of theproposed SocialCircle by training it along with several newly releasedtrajectory prediction models, and experiments show that the SocialCircle notonly quantitatively improves the prediction performance, but also qualitativelyhelps better consider social interactions when forecasting pedestriantrajectories in a way that is consistent with human intuitions.

SocialCircle: Learning the Angle-based Social Interaction Representation for Pedestrian Trajectory Prediction