A hierarchical framework learns generalisable coupling terms for bounded reactive obstacle avoidance by unifying perception, decision and action via low-dimensional geometric descriptors and dynamic movement primitives.
Learning sensor feedback models from demonstrations via phase-modulated neural networks,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Learning Generalisable Coupling Terms for Obstacle Avoidance via Low-dimensional Geometric Descriptors
A hierarchical framework learns generalisable coupling terms for bounded reactive obstacle avoidance by unifying perception, decision and action via low-dimensional geometric descriptors and dynamic movement primitives.