Matches in SemOpenAlex for { <https://semopenalex.org/work/W3206592106> ?p ?o ?g. }
- W3206592106 abstract "Parameterized movement primitives have been extensively used for imitation learning of robotic tasks. However, the high-dimensionality of the parameter space hinders the improvement of such primitives in the reinforcement learning (RL) setting, especially for learning with physical robots. In this paper we propose a novel view on handling the demonstrated trajectories for acquiring low-dimensional, non-linear latent dynamics, using mixtures of probabilistic principal component analyzers (MPPCA) on the movements’ parameter space. Moreover, we introduce a new contextual off-policy RL algorithm, named LAtent-Movements Policy Optimization (LAMPO). LAMPO can provide gradient estimates from previous experience using self-normalized importance sampling, hence, making full use of samples collected in previous learning iterations. These advantages combined provide a complete framework for sample-efficient off-policy optimization of movement primitives for robot learning of high-dimensional manipulation skills. Our experimental results conducted both in simulation and on a real robot show that LAMPO provides sample-efficient policies against common approaches in literature. Code available at https://github.com/SamuelePolimi/lampo." @default.
- W3206592106 created "2021-10-25" @default.
- W3206592106 creator A5026055366 @default.
- W3206592106 creator A5051476305 @default.
- W3206592106 creator A5071367253 @default.
- W3206592106 date "2021-05-30" @default.
- W3206592106 modified "2023-10-13" @default.
- W3206592106 title "Contextual Latent-Movements Off-Policy Optimization for Robotic Manipulation Skills" @default.
- W3206592106 cites W1499669280 @default.
- W3206592106 cites W1513084430 @default.
- W3206592106 cites W1792818999 @default.
- W3206592106 cites W1929309940 @default.
- W3206592106 cites W1968876723 @default.
- W3206592106 cites W1990470929 @default.
- W3206592106 cites W2021004298 @default.
- W3206592106 cites W2042882799 @default.
- W3206592106 cites W2051453180 @default.
- W3206592106 cites W2071444114 @default.
- W3206592106 cites W2116226448 @default.
- W3206592106 cites W2119717200 @default.
- W3206592106 cites W2128677288 @default.
- W3206592106 cites W2134447392 @default.
- W3206592106 cites W2136719407 @default.
- W3206592106 cites W2146610201 @default.
- W3206592106 cites W2213467466 @default.
- W3206592106 cites W2569181534 @default.
- W3206592106 cites W2735021678 @default.
- W3206592106 cites W2802700851 @default.
- W3206592106 cites W2884285175 @default.
- W3206592106 cites W2897555104 @default.
- W3206592106 cites W2994446013 @default.
- W3206592106 cites W3007769740 @default.
- W3206592106 cites W3091497200 @default.
- W3206592106 cites W3101754927 @default.
- W3206592106 doi "https://doi.org/10.1109/icra48506.2021.9561870" @default.
- W3206592106 hasPublicationYear "2021" @default.
- W3206592106 type Work @default.
- W3206592106 sameAs 3206592106 @default.
- W3206592106 citedByCount "4" @default.
- W3206592106 countsByYear W32065921062022 @default.
- W3206592106 countsByYear W32065921062023 @default.
- W3206592106 crossrefType "proceedings-article" @default.
- W3206592106 hasAuthorship W3206592106A5026055366 @default.
- W3206592106 hasAuthorship W3206592106A5051476305 @default.
- W3206592106 hasAuthorship W3206592106A5071367253 @default.
- W3206592106 hasBestOaLocation W32065921062 @default.
- W3206592106 hasConcept C111030470 @default.
- W3206592106 hasConcept C11413529 @default.
- W3206592106 hasConcept C119857082 @default.
- W3206592106 hasConcept C121332964 @default.
- W3206592106 hasConcept C1276947 @default.
- W3206592106 hasConcept C13662910 @default.
- W3206592106 hasConcept C154945302 @default.
- W3206592106 hasConcept C165464430 @default.
- W3206592106 hasConcept C168167062 @default.
- W3206592106 hasConcept C177264268 @default.
- W3206592106 hasConcept C185592680 @default.
- W3206592106 hasConcept C198531522 @default.
- W3206592106 hasConcept C199360897 @default.
- W3206592106 hasConcept C2776760102 @default.
- W3206592106 hasConcept C2778445095 @default.
- W3206592106 hasConcept C41008148 @default.
- W3206592106 hasConcept C43617362 @default.
- W3206592106 hasConcept C49937458 @default.
- W3206592106 hasConcept C90509273 @default.
- W3206592106 hasConcept C97355855 @default.
- W3206592106 hasConcept C97541855 @default.
- W3206592106 hasConceptScore W3206592106C111030470 @default.
- W3206592106 hasConceptScore W3206592106C11413529 @default.
- W3206592106 hasConceptScore W3206592106C119857082 @default.
- W3206592106 hasConceptScore W3206592106C121332964 @default.
- W3206592106 hasConceptScore W3206592106C1276947 @default.
- W3206592106 hasConceptScore W3206592106C13662910 @default.
- W3206592106 hasConceptScore W3206592106C154945302 @default.
- W3206592106 hasConceptScore W3206592106C165464430 @default.
- W3206592106 hasConceptScore W3206592106C168167062 @default.
- W3206592106 hasConceptScore W3206592106C177264268 @default.
- W3206592106 hasConceptScore W3206592106C185592680 @default.
- W3206592106 hasConceptScore W3206592106C198531522 @default.
- W3206592106 hasConceptScore W3206592106C199360897 @default.
- W3206592106 hasConceptScore W3206592106C2776760102 @default.
- W3206592106 hasConceptScore W3206592106C2778445095 @default.
- W3206592106 hasConceptScore W3206592106C41008148 @default.
- W3206592106 hasConceptScore W3206592106C43617362 @default.
- W3206592106 hasConceptScore W3206592106C49937458 @default.
- W3206592106 hasConceptScore W3206592106C90509273 @default.
- W3206592106 hasConceptScore W3206592106C97355855 @default.
- W3206592106 hasConceptScore W3206592106C97541855 @default.
- W3206592106 hasLocation W32065921061 @default.
- W3206592106 hasLocation W32065921062 @default.
- W3206592106 hasOpenAccess W3206592106 @default.
- W3206592106 hasPrimaryLocation W32065921061 @default.
- W3206592106 hasRelatedWork W1494268238 @default.
- W3206592106 hasRelatedWork W154868527 @default.
- W3206592106 hasRelatedWork W1983207144 @default.
- W3206592106 hasRelatedWork W2051058708 @default.
- W3206592106 hasRelatedWork W2116157560 @default.
- W3206592106 hasRelatedWork W2877093712 @default.
- W3206592106 hasRelatedWork W4310614650 @default.
- W3206592106 hasRelatedWork W4367369879 @default.