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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Teaching robots to learn diverse locomotion skills under complexthree-dimensional environmental settings via Reinforcement Learning (RL) isstill challenging. It has been shown that training agents in simple settingsbefore moving them on to complex settings improves the training process, but sofar only in the context of relatively simple locomotion skills. In this work,we adapt the Enhanced Paired Open-Ended Trailblazer (ePOET) approach to trainmore complex agents to walk efficiently on complex three-dimensional terrains.First, to generate more rugged and diverse three-dimensional training terrainswith increasing complexity, we extend the Compositional Pattern ProducingNetworks - Neuroevolution of Augmenting Topologies (CPPN-NEAT) approach andinclude randomized shapes. Second, we combine ePOET with Soft Actor-Criticoff-policy optimization, yielding ePOET-SAC, to ensure that the agent couldlearn more diverse skills to solve more challenging tasks. Our experimentalresults show that the newly generated three-dimensional terrains havesufficient diversity and complexity to guide learning, that ePOET successfullylearns complex locomotion skills on these terrains, and that our proposedePOET-SAC approach slightly improves upon ePOET.

Open-Ended Learning Strategies for Learning Complex Locomotion Skills