DC Field | Value | Language |
---|---|---|
dc.contributor.author | Ha, Jung-Su | ko |
dc.contributor.author | Park, Young-Jin | ko |
dc.contributor.author | Chae, Hyeok-Joo | ko |
dc.contributor.author | Park, Soon-Seo | ko |
dc.contributor.author | Choi, Han-Lim | ko |
dc.date.accessioned | 2023-09-07T01:01:58Z | - |
dc.date.available | 2023-09-07T01:01:58Z | - |
dc.date.created | 2023-09-07 | - |
dc.date.created | 2023-09-07 | - |
dc.date.issued | 2021-05-30 | - |
dc.identifier.citation | 2021 IEEE International Conference on Robotics and Automation (ICRA), pp.4459 - 4466 | - |
dc.identifier.issn | 1050-4729 | - |
dc.identifier.uri | http://hdl.handle.net/10203/312282 | - |
dc.description.abstract | We present a hierarchical planning and control framework that enables an agent to perform various tasks and adapt to a new task flexibly. Rather than learning an individual policy for each particular task, the proposed framework, DISH, distills a hierarchical policy from a set of tasks by representation and reinforcement learning. The framework is based on the idea of latent variable models that represent high-dimensional observations using low-dimensional latent variables. The resulting policy consists of two levels of hierarchy: (i) a planning module that reasons a sequence of latent intentions that would lead to an optimistic future and (ii) a feedback control policy, shared across the tasks, that executes the inferred intention. Because the planning is performed in low-dimensional latent space, the learned policy can immediately be used to solve or adapt to new tasks without additional training. We demonstrate the proposed framework can learn compact representations (3- and 1-dimensional latent states and commands for a humanoid with 197- and 36-dimensional state features and actions) while solving a small number of imitation tasks, and the resulting policy is directly applicable to other types of tasks, i.e., navigation in cluttered environments. | - |
dc.language | English | - |
dc.publisher | IEEE | - |
dc.title | Distilling a Hierarchical Policy for Planning and Control via Representation and Reinforcement Learning | - |
dc.type | Conference | - |
dc.identifier.wosid | 000765738803071 | - |
dc.identifier.scopusid | 2-s2.0-85125471487 | - |
dc.type.rims | CONF | - |
dc.citation.beginningpage | 4459 | - |
dc.citation.endingpage | 4466 | - |
dc.citation.publicationname | 2021 IEEE International Conference on Robotics and Automation (ICRA) | - |
dc.identifier.conferencecountry | CC | - |
dc.identifier.conferencelocation | Xi'an | - |
dc.identifier.doi | 10.1109/icra48506.2021.9561017 | - |
dc.contributor.localauthor | Choi, Han-Lim | - |
dc.contributor.nonIdAuthor | Ha, Jung-Su | - |
dc.contributor.nonIdAuthor | Park, Young-Jin | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.