Matches in SemOpenAlex for { <https://semopenalex.org/work/W3138650132> ?p ?o ?g. }
- W3138650132 abstract "Application of Deep Reinforcement Learning (DRL) algorithms in robotic tasks faces many challenges. On the one hand, reward-shaping for complex tasks that involve multiple sequences is difficult and may result in sub-optimal performances. On the other hand, a sparse-reward setting renders exploration inefficient, and exploration using physical robots is of high-cost and unsafe. In this paper we propose a method of learning long-horizon sparse-reward tasks utilizing one or more existing controllers. Built upon Deep Deterministic Policy Gradients (DDPG), our algorithm incorporates the controllers into stages of exploration, policy update, and most importantly, learning a heuristic value function that naturally interpolates along task trajectories. Through experiments ranging from stacking blocks to cups, we present a straightforward way of synthesizing these controllers, and show that the learned state-based or image-based policies steadily outperform them. Compared to previous works of learning from demonstrations, our method improves sample efficiency by orders of magnitude. Overall, our method bears the potential of leveraging existing industrial robot manipulation systems to build more flexible and intelligent controllers." @default.
- W3138650132 created "2021-03-29" @default.
- W3138650132 creator A5012669482 @default.
- W3138650132 creator A5054513966 @default.
- W3138650132 creator A5055403465 @default.
- W3138650132 creator A5091161457 @default.
- W3138650132 date "2020-11-24" @default.
- W3138650132 modified "2023-09-23" @default.
- W3138650132 title "Achieving Sample-Efficient Learning of Long-Horizon Sparse-Reward Robotic Tasks with Base Controllers" @default.
- W3138650132 cites W2061562262 @default.
- W3138650132 cites W2104733512 @default.
- W3138650132 cites W2121615981 @default.
- W3138650132 cites W2145339207 @default.
- W3138650132 cites W2167224731 @default.
- W3138650132 cites W2173248099 @default.
- W3138650132 cites W2296673577 @default.
- W3138650132 cites W2342840547 @default.
- W3138650132 cites W2575705757 @default.
- W3138650132 cites W2736601468 @default.
- W3138650132 cites W2741122588 @default.
- W3138650132 cites W2785962646 @default.
- W3138650132 cites W2787938642 @default.
- W3138650132 cites W2795561664 @default.
- W3138650132 cites W2810785043 @default.
- W3138650132 cites W2883896749 @default.
- W3138650132 cites W2904246096 @default.
- W3138650132 cites W2952991114 @default.
- W3138650132 cites W2958300349 @default.
- W3138650132 cites W2962793652 @default.
- W3138650132 cites W2962957031 @default.
- W3138650132 cites W2963099939 @default.
- W3138650132 cites W2963277051 @default.
- W3138650132 cites W2964001908 @default.
- W3138650132 cites W2967355195 @default.
- W3138650132 cites W2967727187 @default.
- W3138650132 cites W2968268581 @default.
- W3138650132 cites W2970377754 @default.
- W3138650132 cites W2981030070 @default.
- W3138650132 cites W3008441114 @default.
- W3138650132 cites W3079699067 @default.
- W3138650132 cites W3094336301 @default.
- W3138650132 cites W3130717831 @default.
- W3138650132 hasPublicationYear "2020" @default.
- W3138650132 type Work @default.
- W3138650132 sameAs 3138650132 @default.
- W3138650132 citedByCount "0" @default.
- W3138650132 crossrefType "posted-content" @default.
- W3138650132 hasAuthorship W3138650132A5012669482 @default.
- W3138650132 hasAuthorship W3138650132A5054513966 @default.
- W3138650132 hasAuthorship W3138650132A5055403465 @default.
- W3138650132 hasAuthorship W3138650132A5091161457 @default.
- W3138650132 hasConcept C119857082 @default.
- W3138650132 hasConcept C127413603 @default.
- W3138650132 hasConcept C14036430 @default.
- W3138650132 hasConcept C154945302 @default.
- W3138650132 hasConcept C173801870 @default.
- W3138650132 hasConcept C185592680 @default.
- W3138650132 hasConcept C196340769 @default.
- W3138650132 hasConcept C198531522 @default.
- W3138650132 hasConcept C201995342 @default.
- W3138650132 hasConcept C2780451532 @default.
- W3138650132 hasConcept C41008148 @default.
- W3138650132 hasConcept C43617362 @default.
- W3138650132 hasConcept C78458016 @default.
- W3138650132 hasConcept C86803240 @default.
- W3138650132 hasConcept C90509273 @default.
- W3138650132 hasConcept C97541855 @default.
- W3138650132 hasConceptScore W3138650132C119857082 @default.
- W3138650132 hasConceptScore W3138650132C127413603 @default.
- W3138650132 hasConceptScore W3138650132C14036430 @default.
- W3138650132 hasConceptScore W3138650132C154945302 @default.
- W3138650132 hasConceptScore W3138650132C173801870 @default.
- W3138650132 hasConceptScore W3138650132C185592680 @default.
- W3138650132 hasConceptScore W3138650132C196340769 @default.
- W3138650132 hasConceptScore W3138650132C198531522 @default.
- W3138650132 hasConceptScore W3138650132C201995342 @default.
- W3138650132 hasConceptScore W3138650132C2780451532 @default.
- W3138650132 hasConceptScore W3138650132C41008148 @default.
- W3138650132 hasConceptScore W3138650132C43617362 @default.
- W3138650132 hasConceptScore W3138650132C78458016 @default.
- W3138650132 hasConceptScore W3138650132C86803240 @default.
- W3138650132 hasConceptScore W3138650132C90509273 @default.
- W3138650132 hasConceptScore W3138650132C97541855 @default.
- W3138650132 hasLocation W31386501321 @default.
- W3138650132 hasOpenAccess W3138650132 @default.
- W3138650132 hasPrimaryLocation W31386501321 @default.
- W3138650132 hasRelatedWork W2528734395 @default.
- W3138650132 hasRelatedWork W2554830522 @default.
- W3138650132 hasRelatedWork W2789102901 @default.
- W3138650132 hasRelatedWork W2903497309 @default.
- W3138650132 hasRelatedWork W2907537824 @default.
- W3138650132 hasRelatedWork W2950622182 @default.
- W3138650132 hasRelatedWork W2953082648 @default.
- W3138650132 hasRelatedWork W2953129341 @default.
- W3138650132 hasRelatedWork W2966950754 @default.
- W3138650132 hasRelatedWork W2970720334 @default.
- W3138650132 hasRelatedWork W2991420892 @default.
- W3138650132 hasRelatedWork W2996006200 @default.
- W3138650132 hasRelatedWork W3045280543 @default.
- W3138650132 hasRelatedWork W3046387304 @default.