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

Dynamic quadruped locomotion over challenging terrains with precise footplacements is a hard problem for both optimal control methods and ReinforcementLearning (RL). Non-linear solvers can produce coordinated constraint satisfyingmotions, but often take too long to converge for online application. RL methodscan learn dynamic reactive controllers but require carefully tuned shapingrewards to produce good gaits and can have trouble discovering precisecoordinated movements. Imitation learning circumvents this problem and has beenused with motion capture data to extract quadruped gaits for flat terrains.However, it would be costly to acquire motion capture data for a very largevariety of terrains with height differences. In this work, we combine theadvantages of trajectory optimization and learning methods and show thatterrain adaptive controllers can be obtained by training policies to imitatetrajectories that have been planned over procedural terrains by a non-linearsolver. We show that the learned policies transfer to unseen terrains and canbe fine-tuned to dynamically traverse challenging terrains that require precisefoot placements and are very hard to solve with standard RL.

Learning Coordinated Terrain-Adaptive Locomotion by Imitating a Centroidal Dynamics Planner