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 complex environments without a global map is along-standing challenge for mobile robots. While deep reinforcement learning(DRL) has attracted a rapidly growing interest in solving such an autonomousnavigation problem for its generalization capability, DRL typically leads to amediocre navigation performance in practice due to the gap between the trainingscene and the actual test scene. Most existing work focuses on tuning thealgorithm to enhance its transferability, whereas few investigates how toquantify or measure the gap therebetween. This paper proposes a newtransferability metric -- the scene similarity calculated using an improvedimage template matching algorithm to quantify the transferability of a DRLagent between the training and test scenes. Specifically, two transferabilityperformance indicators are designed including the global scene similarity thatevaluates the overall robustness of a DRL algorithm and the local scenesimilarity that serves as a safety measure when a DRL agent is deployed withouta global map. In addition, this paper proposes the use of a local map thatfuses 2D LiDAR data with spatial information of both the agent and thedestination as the DRL observation, aiming to improve the transferability ofDRL navigation algorithms. With a wheeled robot as the case study platform,both simulation and real-world experiments are conducted in a total of 26different scenes. The experimental results confirm the robustness of the localmap observation design and validate the scene similarity metric in predictingthe success rate of DRL navigation algorithms.

A Tansferability Metric Using Scene Similarity and Local Map Observation for DRL Navigation