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 navigation in off-road conditions requires an accurate estimationof terrain traversability. However, traversability estimation in unstructuredenvironments is subject to high uncertainty due to the variability of numerousfactors that influence vehicle-terrain interaction. Consequently, it ischallenging to obtain a generalizable model that can accurately predicttraversability in a variety of environments. This paper presents METAVerse, ameta-learning framework for learning a global model that accurately andreliably predicts terrain traversability across diverse environments. We trainthe traversability prediction network to generate a dense and continuous-valuedcost map from a sparse LiDAR point cloud, leveraging vehicle-terraininteraction feedback in a self-supervised manner. Meta-learning is utilized totrain a global model with driving data collected from multiple environments,effectively minimizing estimation uncertainty. During deployment, onlineadaptation is performed to rapidly adapt the network to the local environmentby exploiting recent interaction experiences. To conduct a comprehensiveevaluation, we collect driving data from various terrains and demonstrate thatour method can obtain a global model that minimizes uncertainty. Moreover, byintegrating our model with a model predictive controller, we demonstrate thatthe reduced uncertainty results in safe and stable navigation in unstructuredand unknown terrains.

METAVerse: Meta-Learning Traversability Cost Map for Off-Road Navigation