Matches in SemOpenAlex for { <https://semopenalex.org/work/W4383108478> ?p ?o ?g. }
- W4383108478 abstract "In-hand manipulation is challenging for a multi-finger robotic hand due to its high degrees of freedom and the complex interaction with the object. To enable in-hand manipulation, existing deep reinforcement learning based approaches mainly focus on training a single robot-structure-specific policy through the centralized learning mechanism, lacking adaptability to changes like robot malfunction. To solve this limitation, this work treats each finger as an individual agent and trains multiple agents to control their assigned fingers to complete the in-hand manipulation task cooperatively. We propose the Multi-Agent Global-Observation Critic and Local-Observation Actor (MAGCLA) method, where the critic can observe all agents' actions globally, and the actor only locally observes its neighbors' actions. Besides, conventional individual experience replay may cause unstable cooperation due to the asynchronous performance increment of each agent, which is critical for in-hand manipulation tasks. To solve this issue, we propose the Synchronized Hindsight Experience Replay (SHER) method to synchronize and efficiently reuse the replayed experience across all agents. The methods are evaluated in two in-hand manipulation tasks on the Shadow dexterous hand. The results show that SHER helps MAGCLA achieve comparable learning efficiency to a single policy, and the MAGCLA approach is more generalizable in different tasks. The trained policies have higher adaptability in the robot malfunction test compared to the baseline multi-agent and single-agent approaches." @default.
- W4383108478 created "2023-07-05" @default.
- W4383108478 creator A5012912557 @default.
- W4383108478 creator A5014940997 @default.
- W4383108478 creator A5054228072 @default.
- W4383108478 creator A5079248597 @default.
- W4383108478 date "2023-05-29" @default.
- W4383108478 modified "2023-09-24" @default.
- W4383108478 title "A Multi-Agent Approach for Adaptive Finger Cooperation in Learning-based In-Hand Manipulation" @default.
- W4383108478 cites W1542941925 @default.
- W4383108478 cites W1974706266 @default.
- W4383108478 cites W2101008389 @default.
- W4383108478 cites W2154740693 @default.
- W4383108478 cites W2158782408 @default.
- W4383108478 cites W2171879251 @default.
- W4383108478 cites W2617547828 @default.
- W4383108478 cites W2774971480 @default.
- W4383108478 cites W2887316838 @default.
- W4383108478 cites W2962680503 @default.
- W4383108478 cites W2990747716 @default.
- W4383108478 cites W3100789280 @default.
- W4383108478 cites W3130418699 @default.
- W4383108478 cites W3132708124 @default.
- W4383108478 cites W3140147451 @default.
- W4383108478 cites W3175827959 @default.
- W4383108478 cites W3179018212 @default.
- W4383108478 cites W3187695488 @default.
- W4383108478 cites W3205469446 @default.
- W4383108478 cites W3210090000 @default.
- W4383108478 cites W4206787406 @default.
- W4383108478 cites W4225372550 @default.
- W4383108478 cites W971303027 @default.
- W4383108478 doi "https://doi.org/10.1109/icra48891.2023.10160909" @default.
- W4383108478 hasPublicationYear "2023" @default.
- W4383108478 type Work @default.
- W4383108478 citedByCount "0" @default.
- W4383108478 crossrefType "proceedings-article" @default.
- W4383108478 hasAuthorship W4383108478A5012912557 @default.
- W4383108478 hasAuthorship W4383108478A5014940997 @default.
- W4383108478 hasAuthorship W4383108478A5054228072 @default.
- W4383108478 hasAuthorship W4383108478A5079248597 @default.
- W4383108478 hasBestOaLocation W43831084782 @default.
- W4383108478 hasConcept C10347200 @default.
- W4383108478 hasConcept C107457646 @default.
- W4383108478 hasConcept C120665830 @default.
- W4383108478 hasConcept C121332964 @default.
- W4383108478 hasConcept C127413603 @default.
- W4383108478 hasConcept C151319957 @default.
- W4383108478 hasConcept C154945302 @default.
- W4383108478 hasConcept C15744967 @default.
- W4383108478 hasConcept C177606310 @default.
- W4383108478 hasConcept C180747234 @default.
- W4383108478 hasConcept C18903297 @default.
- W4383108478 hasConcept C192209626 @default.
- W4383108478 hasConcept C201995342 @default.
- W4383108478 hasConcept C2780451532 @default.
- W4383108478 hasConcept C2781238097 @default.
- W4383108478 hasConcept C31258907 @default.
- W4383108478 hasConcept C41008148 @default.
- W4383108478 hasConcept C58581272 @default.
- W4383108478 hasConcept C86803240 @default.
- W4383108478 hasConcept C90509273 @default.
- W4383108478 hasConcept C97541855 @default.
- W4383108478 hasConceptScore W4383108478C10347200 @default.
- W4383108478 hasConceptScore W4383108478C107457646 @default.
- W4383108478 hasConceptScore W4383108478C120665830 @default.
- W4383108478 hasConceptScore W4383108478C121332964 @default.
- W4383108478 hasConceptScore W4383108478C127413603 @default.
- W4383108478 hasConceptScore W4383108478C151319957 @default.
- W4383108478 hasConceptScore W4383108478C154945302 @default.
- W4383108478 hasConceptScore W4383108478C15744967 @default.
- W4383108478 hasConceptScore W4383108478C177606310 @default.
- W4383108478 hasConceptScore W4383108478C180747234 @default.
- W4383108478 hasConceptScore W4383108478C18903297 @default.
- W4383108478 hasConceptScore W4383108478C192209626 @default.
- W4383108478 hasConceptScore W4383108478C201995342 @default.
- W4383108478 hasConceptScore W4383108478C2780451532 @default.
- W4383108478 hasConceptScore W4383108478C2781238097 @default.
- W4383108478 hasConceptScore W4383108478C31258907 @default.
- W4383108478 hasConceptScore W4383108478C41008148 @default.
- W4383108478 hasConceptScore W4383108478C58581272 @default.
- W4383108478 hasConceptScore W4383108478C86803240 @default.
- W4383108478 hasConceptScore W4383108478C90509273 @default.
- W4383108478 hasConceptScore W4383108478C97541855 @default.
- W4383108478 hasFunder F4320306076 @default.
- W4383108478 hasLocation W43831084781 @default.
- W4383108478 hasLocation W43831084782 @default.
- W4383108478 hasLocation W43831084783 @default.
- W4383108478 hasOpenAccess W4383108478 @default.
- W4383108478 hasPrimaryLocation W43831084781 @default.
- W4383108478 hasRelatedWork W2005616670 @default.
- W4383108478 hasRelatedWork W2016398835 @default.
- W4383108478 hasRelatedWork W2024276883 @default.
- W4383108478 hasRelatedWork W3081510516 @default.
- W4383108478 hasRelatedWork W3168256553 @default.
- W4383108478 hasRelatedWork W3186042173 @default.
- W4383108478 hasRelatedWork W3186584605 @default.
- W4383108478 hasRelatedWork W3199692841 @default.
- W4383108478 hasRelatedWork W4287124629 @default.
- W4383108478 hasRelatedWork W4298128006 @default.