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

Autonomous vehicles (AVs) need to reason about the multimodal behavior ofneighboring agents while planning their own motion. Many existing trajectoryplanners seek a single trajectory that performs well under \emph{all} plausiblefutures simultaneously, ignoring bi-directional interactions and thus leadingto overly conservative plans. Policy planning, whereby the ego agent plans apolicy that reacts to the environment's multimodal behavior, is a promisingdirection as it can account for the action-reaction interactions between the AVand the environment. However, most existing policy planners do not scale to thecomplexity of real autonomous vehicle applications: they are either notcompatible with modern deep learning prediction models, not interpretable, ornot able to generate high quality trajectories. To fill this gap, we proposeTree Policy Planning (TPP), a policy planner that is compatible withstate-of-the-art deep learning prediction models, generates multistage motionplans, and accounts for the influence of ego agent on the environment behavior.The key idea of TPP is to reduce the continuous optimization problem into atractable discrete Markov Decision Process (MDP) through the construction oftwo tree structures: an ego trajectory tree for ego trajectory options, and ascenario tree for multi-modal ego-conditioned environment predictions. Wedemonstrate the efficacy of TPP in closed-loop simulations based on real-worldnuScenes dataset and results show that TPP scales to realistic AV scenarios andsignificantly outperforms non-policy baselines.

Tree-structured Policy Planning with Learned Behavior Models