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

We address key challenges in long-horizon embodied exploration and navigationby proposing a new object transport task and a novel modular framework fortemporally extended navigation. Our first contribution is the design of a novelLong-HOT environment focused on deep exploration and long-horizon planningwhere the agent is required to efficiently find and pick up target objects tobe carried and dropped at a goal location, with load constraints and optionalaccess to a container if it finds one. Further, we propose a modularhierarchical transport policy (HTP) that builds a topological graph of thescene to perform exploration with the help of weighted frontiers. Ourhierarchical approach uses a combination of motion planning algorithms to reachpoint goals within explored locations and object navigation policies for movingtowards semantic targets at unknown locations. Experiments on both our proposedHabitat transport task and on MultiOn benchmarks show that our methodsignificantly outperforms baselines and prior works. Further, we validate theeffectiveness of our modular approach for long-horizon transport bydemonstrating meaningful generalization to much harder transport scenes withtraining only on simpler versions of the task.

Long-HOT: A Modular Hierarchical Approach for Long-Horizon Object Transport