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

While the modeling of pair-wise relations has been widely studied inmulti-agent interacting systems, its ability to capture higher-level andlarger-scale group-wise activities is limited. In this paper, we propose agroup-aware relational reasoning approach (named EvolveHypergraph) withexplicit inference of the underlying dynamically evolving relationalstructures, and we demonstrate its effectiveness for multi-agent trajectoryprediction. In addition to the edges between a pair of nodes (i.e., agents), wepropose to infer hyperedges that adaptively connect multiple nodes to enablegroup-aware relational reasoning in an unsupervised manner without fixing thenumber of hyperedges. The proposed approach infers the dynamically evolvingrelation graphs and hypergraphs over time to capture the evolution ofrelations, which are used by the trajectory predictor to obtain future states.Moreover, we propose to regularize the smoothness of the relation evolution andthe sparsity of the inferred graphs or hypergraphs, which effectively improvestraining stability and enhances the explainability of inferred relations. Theproposed approach is validated on both synthetic crowd simulations and multiplereal-world benchmark datasets. Our approach infers explainable, reasonablegroup-aware relations and achieves state-of-the-art performance in long-termprediction.

EvolveHypergraph: Group-Aware Dynamic Relational Reasoning for Trajectory Prediction