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

Autonomous driving in urban crowds at unregulated intersections ischallenging, where dynamic occlusions and uncertain behaviors of other vehiclesshould be carefully considered. Traditional methods are heuristic and based onhand-engineered rules and parameters, but scale poorly in new situations.Therefore, they require high labor cost to design and maintain rules in allforeseeable scenarios. Recently, deep reinforcement learning (DRL) has shownpromising results in urban driving scenarios. However, DRL is known to besample inefficient, and most previous works assume perfect observations such asground-truth locations and motions of vehicles without considering noises andocclusions, which might be a too strong assumption for policy deployment. Inthis work, we use DRL to train lidar-based end-to-end driving policies thatnaturally consider imperfect partial observations. We further use unsupervisedcontrastive representation learning as an auxiliary task to improve the sampleefficiency. The comparative evaluation results reveal that our method achieveshigher success rates than the state-of-the-art (SOTA) lidar-based end-to-enddriving network, better trades off safety and efficiency than the carefullytuned rule-based method, and generalizes better to new scenarios than thebaselines. Demo videos are available at https://caipeide.github.io/carl-lead/.

Carl-Lead: Lidar-based End-to-End Autonomous Driving with Contrastive Deep Reinforcement Learning