FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3206827162> ?p ?o ?g. }

• W3206827162 abstract "Reinforcement learning (RL) can in principle let robots automatically adapt to new tasks, but current RL methods require a large number of trials to accomplish this. In this paper, we tackle rapid adaptation to new tasks through the framework of meta-learning, which utilizes past tasks to learn to adapt with a specific focus on industrial insertion tasks. Fast adaptation is crucial because prohibitively large number of on-robot trials will potentially damage hardware pieces. Additionally, effective adaptation is also feasible in that experience among different insertion applications can be largely leveraged by each other. In this setting, we address two specific challenges when applying meta-learning. First, conventional meta-RL algorithms require lengthy online meta-training. We show that this can be replaced with appropriately chosen offline data, resulting in an offline meta- RL method that only requires demonstrations and trials from each of the prior tasks, without the need to run costly meta-RL procedures online. Second, meta-RL methods can fail to generalize to new tasks that are too different from those seen at meta-training time, which poses a particular challenge in industrial applications, where high success rates are critical. We address this by combining contextual meta-learning with direct online finetuning: if the new task is similar to those seen in the prior data, then the contextual meta-learner adapts immediately, and if it is too different, it gradually adapts through finetuning. We show that our approach is able to quickly adapt to a variety of different insertion tasks, with a success rate of 100% using only a fraction of the samples needed for learning the tasks from scratch. Experiment videos and details are available at //sites.google.com/view/offline-metarl-insertion.https:" @default.
• W3206827162 created "2021-10-25" @default.
• W3206827162 creator A5026322200 @default.
• W3206827162 creator A5029642293 @default.
• W3206827162 creator A5037464692 @default.
• W3206827162 creator A5041306134 @default.
• W3206827162 creator A5061512729 @default.
• W3206827162 creator A5062951341 @default.
• W3206827162 creator A5063215180 @default.
• W3206827162 creator A5077984643 @default.
• W3206827162 date "2022-05-23" @default.
• W3206827162 modified "2023-09-25" @default.
• W3206827162 title "Offline Meta-Reinforcement Learning for Industrial Insertion" @default.
• W3206827162 cites W1938885652 @default.
• W3206827162 cites W2008731016 @default.
• W3206827162 cites W2034052310 @default.
• W3206827162 cites W2110697446 @default.
• W3206827162 cites W2111067802 @default.
• W3206827162 cites W2115617434 @default.
• W3206827162 cites W2539574638 @default.
• W3206827162 cites W2553722312 @default.
• W3206827162 cites W2962732398 @default.
• W3206827162 cites W2963403593 @default.
• W3206827162 cites W2964333597 @default.
• W3206827162 cites W2967355195 @default.
• W3206827162 cites W2968268581 @default.
• W3206827162 cites W3090369311 @default.
• W3206827162 cites W3130717831 @default.
• W3206827162 cites W3130984490 @default.
• W3206827162 cites W3159735414 @default.
• W3206827162 cites W3174364619 @default.
• W3206827162 doi "https://doi.org/10.1109/icra46639.2022.9812312" @default.
• W3206827162 hasPublicationYear "2022" @default.
• W3206827162 type Work @default.
• W3206827162 sameAs 3206827162 @default.
• W3206827162 citedByCount "9" @default.
• W3206827162 countsByYear W32068271622021 @default.
• W3206827162 countsByYear W32068271622022 @default.
• W3206827162 countsByYear W32068271622023 @default.
• W3206827162 crossrefType "proceedings-article" @default.
• W3206827162 hasAuthorship W3206827162A5026322200 @default.
• W3206827162 hasAuthorship W3206827162A5029642293 @default.
• W3206827162 hasAuthorship W3206827162A5037464692 @default.
• W3206827162 hasAuthorship W3206827162A5041306134 @default.
• W3206827162 hasAuthorship W3206827162A5061512729 @default.
• W3206827162 hasAuthorship W3206827162A5062951341 @default.
• W3206827162 hasAuthorship W3206827162A5063215180 @default.
• W3206827162 hasAuthorship W3206827162A5077984643 @default.
• W3206827162 hasBestOaLocation W32068271622 @default.
• W3206827162 hasConcept C119857082 @default.
• W3206827162 hasConcept C120665830 @default.
• W3206827162 hasConcept C121332964 @default.
• W3206827162 hasConcept C136197465 @default.
• W3206827162 hasConcept C139807058 @default.
• W3206827162 hasConcept C154945302 @default.
• W3206827162 hasConcept C162324750 @default.
• W3206827162 hasConcept C187736073 @default.
• W3206827162 hasConcept C2780451532 @default.
• W3206827162 hasConcept C2781002164 @default.
• W3206827162 hasConcept C41008148 @default.
• W3206827162 hasConcept C90509273 @default.
• W3206827162 hasConcept C97541855 @default.
• W3206827162 hasConceptScore W3206827162C119857082 @default.
• W3206827162 hasConceptScore W3206827162C120665830 @default.
• W3206827162 hasConceptScore W3206827162C121332964 @default.
• W3206827162 hasConceptScore W3206827162C136197465 @default.
• W3206827162 hasConceptScore W3206827162C139807058 @default.
• W3206827162 hasConceptScore W3206827162C154945302 @default.
• W3206827162 hasConceptScore W3206827162C162324750 @default.
• W3206827162 hasConceptScore W3206827162C187736073 @default.
• W3206827162 hasConceptScore W3206827162C2780451532 @default.
• W3206827162 hasConceptScore W3206827162C2781002164 @default.
• W3206827162 hasConceptScore W3206827162C41008148 @default.
• W3206827162 hasConceptScore W3206827162C90509273 @default.
• W3206827162 hasConceptScore W3206827162C97541855 @default.
• W3206827162 hasLocation W32068271621 @default.
• W3206827162 hasLocation W32068271622 @default.
• W3206827162 hasOpenAccess W3206827162 @default.
• W3206827162 hasPrimaryLocation W32068271621 @default.
• W3206827162 hasRelatedWork W2906917062 @default.
• W3206827162 hasRelatedWork W3037114224 @default.
• W3206827162 hasRelatedWork W3090436287 @default.
• W3206827162 hasRelatedWork W3161250517 @default.
• W3206827162 hasRelatedWork W3199608561 @default.
• W3206827162 hasRelatedWork W4226029999 @default.
• W3206827162 hasRelatedWork W4287647350 @default.
• W3206827162 hasRelatedWork W4319083788 @default.
• W3206827162 hasRelatedWork W4319309271 @default.
• W3206827162 hasRelatedWork W4379251374 @default.
• W3206827162 isParatext "false" @default.
• W3206827162 isRetracted "false" @default.
• W3206827162 magId "3206827162" @default.
• W3206827162 workType "article" @default.