Matches in SemOpenAlex for { <https://semopenalex.org/work/W4320854591> ?p ?o ?g. }
Showing items 1 to 59 of
59
with 100 items per page.
- W4320854591 abstract "Adversarial imitation learning has become a widely used imitation learning framework. The discriminator is often trained by taking expert demonstrations and policy trajectories as examples respectively from two categories (positive vs. negative) and the policy is then expected to produce trajectories that are indistinguishable from the expert demonstrations. But in the real world, the collected expert demonstrations are more likely to be imperfect, where only an unknown fraction of the demonstrations are optimal. Instead of treating imperfect expert demonstrations as absolutely positive or negative, we investigate unlabeled imperfect expert demonstrations as they are. A positive-unlabeled adversarial imitation learning algorithm is developed to dynamically sample expert demonstrations that can well match the trajectories from the constantly optimized agent policy. The trajectories of an initial agent policy could be closer to those non-optimal expert demonstrations, but within the framework of adversarial imitation learning, agent policy will be optimized to cheat the discriminator and produce trajectories that are similar to those optimal expert demonstrations. Theoretical analysis shows that our method learns from the imperfect demonstrations via a self-paced way. Experimental results on MuJoCo and RoboSuite platforms demonstrate the effectiveness of our method from different aspects." @default.
- W4320854591 created "2023-02-16" @default.
- W4320854591 creator A5042785211 @default.
- W4320854591 creator A5071287470 @default.
- W4320854591 creator A5077062257 @default.
- W4320854591 date "2023-02-13" @default.
- W4320854591 modified "2023-09-28" @default.
- W4320854591 title "Unlabeled Imperfect Demonstrations in Adversarial Imitation Learning" @default.
- W4320854591 doi "https://doi.org/10.48550/arxiv.2302.06271" @default.
- W4320854591 hasPublicationYear "2023" @default.
- W4320854591 type Work @default.
- W4320854591 citedByCount "0" @default.
- W4320854591 crossrefType "posted-content" @default.
- W4320854591 hasAuthorship W4320854591A5042785211 @default.
- W4320854591 hasAuthorship W4320854591A5071287470 @default.
- W4320854591 hasAuthorship W4320854591A5077062257 @default.
- W4320854591 hasBestOaLocation W43208545911 @default.
- W4320854591 hasConcept C119857082 @default.
- W4320854591 hasConcept C126388530 @default.
- W4320854591 hasConcept C138885662 @default.
- W4320854591 hasConcept C154945302 @default.
- W4320854591 hasConcept C15744967 @default.
- W4320854591 hasConcept C2779803651 @default.
- W4320854591 hasConcept C2780310539 @default.
- W4320854591 hasConcept C37736160 @default.
- W4320854591 hasConcept C41008148 @default.
- W4320854591 hasConcept C41895202 @default.
- W4320854591 hasConcept C76155785 @default.
- W4320854591 hasConcept C77805123 @default.
- W4320854591 hasConcept C94915269 @default.
- W4320854591 hasConceptScore W4320854591C119857082 @default.
- W4320854591 hasConceptScore W4320854591C126388530 @default.
- W4320854591 hasConceptScore W4320854591C138885662 @default.
- W4320854591 hasConceptScore W4320854591C154945302 @default.
- W4320854591 hasConceptScore W4320854591C15744967 @default.
- W4320854591 hasConceptScore W4320854591C2779803651 @default.
- W4320854591 hasConceptScore W4320854591C2780310539 @default.
- W4320854591 hasConceptScore W4320854591C37736160 @default.
- W4320854591 hasConceptScore W4320854591C41008148 @default.
- W4320854591 hasConceptScore W4320854591C41895202 @default.
- W4320854591 hasConceptScore W4320854591C76155785 @default.
- W4320854591 hasConceptScore W4320854591C77805123 @default.
- W4320854591 hasConceptScore W4320854591C94915269 @default.
- W4320854591 hasLocation W43208545911 @default.
- W4320854591 hasOpenAccess W4320854591 @default.
- W4320854591 hasPrimaryLocation W43208545911 @default.
- W4320854591 hasRelatedWork W2952541330 @default.
- W4320854591 hasRelatedWork W3004128202 @default.
- W4320854591 hasRelatedWork W3037598048 @default.
- W4320854591 hasRelatedWork W3102502161 @default.
- W4320854591 hasRelatedWork W3152795989 @default.
- W4320854591 hasRelatedWork W3200833855 @default.
- W4320854591 hasRelatedWork W4287871966 @default.
- W4320854591 hasRelatedWork W4297631487 @default.
- W4320854591 hasRelatedWork W4287209769 @default.
- W4320854591 hasRelatedWork W4287686113 @default.
- W4320854591 isParatext "false" @default.
- W4320854591 isRetracted "false" @default.
- W4320854591 workType "article" @default.