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 propose a structured prediction approach for robot imitation learning fromdemonstrations. Among various tools for robot imitation learning, supervisedlearning has been observed to have a prominent role. Structured prediction is aform of supervised learning that enables learning models to operate on outputspaces with complex structures. Through the lens of structured prediction, weshow how robots can learn to imitate trajectories belonging to not onlyEuclidean spaces but also Riemannian manifolds. Exploiting ideas frominformation theory, we propose a class of loss functions based on thef-divergence to measure the information loss between the demonstrated andreproduced probabilistic trajectories. Different types of f-divergence willresult in different policies, which we call imitation modes. Furthermore, ourapproach enables the incorporation of spatial and temporal trajectorymodulation, which is necessary for robots to be adaptive to the change inworking conditions. We benchmark our algorithm against state-of-the-art methodsin terms of trajectory reproduction and adaptation. The quantitative evaluationshows that our approach outperforms other algorithms regarding both accuracyand efficiency. We also report real-world experimental results on learningmanifold trajectories in a polishing task with a KUKA LWR robot arm,illustrating the effectiveness of our algorithmic framework.

A Structured Prediction Approach for Robot Imitation Learning