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

Making safe and human-like decisions is an essential capability of autonomousdriving systems, and learning-based behavior planning presents a promisingpathway toward achieving this objective. Distinguished from existinglearning-based methods that directly output decisions, this work introduces apredictive behavior planning framework that learns to predict and evaluate fromhuman driving data. This framework consists of three components: a behaviorgeneration module that produces a diverse set of candidate behaviors in theform of trajectory proposals, a conditional motion prediction network thatpredicts future trajectories of other agents based on each proposal, and ascoring module that evaluates the candidate plans using maximum entropy inversereinforcement learning (IRL). We validate the proposed framework on alarge-scale real-world urban driving dataset through comprehensive experiments.The results show that the conditional prediction model can predict distinct andreasonable future trajectories given different trajectory proposals and theIRL-based scoring module can select plans that are close to human driving. Theproposed framework outperforms other baseline methods in terms of similarity tohuman driving trajectories. Additionally, we find that the conditionalprediction model improves both prediction and planning performance compared tothe non-conditional model. Lastly, we note that learning the scoring module iscrucial for aligning the evaluations with human drivers.

Conditional Predictive Behavior Planning with Inverse Reinforcement Learning for Human-like Autonomous Driving