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

Accurate localization is a critical requirement for most robotic tasks. Themain body of existing work is focused on passive localization in which themotions of the robot are assumed given, abstracting from their influence onsampling informative observations. While recent work has shown the benefits oflearning motions to disambiguate the robot's poses, these methods arerestricted to granular discrete actions and directly depend on the size of theglobal map. We propose Active Particle Filter Networks (APFN), an approach thatonly relies on local information for both the likelihood evaluation as well asthe decision making. To do so, we couple differentiable particle filters with areinforcement learning agent that attends to the most relevant parts of themap. The resulting approach inherits the computational benefits of particlefilters and can directly act in continuous action spaces while remaining fullydifferentiable and thereby end-to-end optimizable as well as agnostic to theinput modality. We demonstrate the benefits of our approach with extensiveexperiments in photorealistic indoor environments built from real-world 3Dscanned apartments. Videos and code are available athttp://apfn.cs.uni-freiburg.de.

Active Particle Filter Networks: Efficient Active Localization in Continuous Action Spaces and Large Maps