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

Trajectory prediction in urban mixed-traffic zones (a.k.a. shared spaces) iscritical for many intelligent transportation systems, such as intent detectionfor autonomous driving. However, there are many challenges to predict thetrajectories of heterogeneous road agents (pedestrians, cyclists and vehicles)at a microscopical level. For example, an agent might be able to choosemultiple plausible paths in complex interactions with other agents in varyingenvironments. To this end, we propose an approach named Multi-Context EncoderNetwork (MCENET) that is trained by encoding both past and future scenecontext, interaction context and motion information to capture the patterns andvariations of the future trajectories using a set of stochastic latentvariables. In inference time, we combine the past context and motioninformation of the target agent with samplings of the latent variables topredict multiple realistic trajectories in the future. Through experiments onseveral datasets of varying scenes, our method outperforms some of the recentstate-of-the-art methods for mixed traffic trajectory prediction by a largemargin and more robust in a very challenging environment. The impact of eachcontext is justified via ablation studies.

MCENET: Multi-Context Encoder Network for Homogeneous Agent Trajectory Prediction in Mixed Traffic