Multi-Goal Reinforcement Learning: Challenging Robotics Environments and Request for Research
The purpose of this technical report is two-fold. First of all, it introduces a suite of challenging continuous control tasks (integrated with OpenAI Gym) based on currently existing robotics hardware. The tasks include pushing, sliding and pick & place with a Fetch robotic arm as well as in-han...
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , |
|---|---|
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | |
|---|---|
| container_issue | |
| container_start_page | |
| container_title | |
| container_volume | |
| creator | Plappert, Matthias Andrychowicz, Marcin Ray, Alex McGrew, Bob Baker, Bowen Powell, Glenn Schneider, Jonas Tobin, Josh Chociej, Maciek Welinder, Peter Kumar, Vikash Zaremba, Wojciech |
| description | The purpose of this technical report is two-fold. First of all, it introduces
a suite of challenging continuous control tasks (integrated with OpenAI Gym)
based on currently existing robotics hardware. The tasks include pushing,
sliding and pick & place with a Fetch robotic arm as well as in-hand object
manipulation with a Shadow Dexterous Hand. All tasks have sparse binary rewards
and follow a Multi-Goal Reinforcement Learning (RL) framework in which an agent
is told what to do using an additional input.
The second part of the paper presents a set of concrete research ideas for
improving RL algorithms, most of which are related to Multi-Goal RL and
Hindsight Experience Replay. |
| doi_str_mv | 10.48550/arxiv.1802.09464 |
| format | Article |
| fullrecord | <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_1802_09464</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>1802_09464</sourcerecordid><originalsourceid>FETCH-LOGICAL-a1154-3215e109ff18781f686f3c199756865e8bfa639e36f071ddef43082bc7173f4c3</originalsourceid><addsrcrecordid>eNotj0FOwzAURL1hgQoHYFVfIME_thOHHYpKQQpCqugSRY7z3VpKbXDSCm6PW1jNjDQz0iPkDlgulJTsXsdvd8pBsSJntSjFNfl4PY6zy9ZBj3SDztsQDR7Qz7RFHb3zuwfa7PU4ot-lQDehD7MzE135k4vBn6sT1X5I668jTjNND8lPaW32N-TK6nHC239dkO3T6r15ztq39Uvz2GYaQIqMFyARWG0tqEqBLVVpuYG6rmSyElVvdclr5KVlFQwDWsGZKnpTQcWtMHxBln-_F8DuM7qDjj_dGbS7gPJfGpdOSQ</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Multi-Goal Reinforcement Learning: Challenging Robotics Environments and Request for Research</title><source>arXiv.org</source><creator>Plappert, Matthias ; Andrychowicz, Marcin ; Ray, Alex ; McGrew, Bob ; Baker, Bowen ; Powell, Glenn ; Schneider, Jonas ; Tobin, Josh ; Chociej, Maciek ; Welinder, Peter ; Kumar, Vikash ; Zaremba, Wojciech</creator><creatorcontrib>Plappert, Matthias ; Andrychowicz, Marcin ; Ray, Alex ; McGrew, Bob ; Baker, Bowen ; Powell, Glenn ; Schneider, Jonas ; Tobin, Josh ; Chociej, Maciek ; Welinder, Peter ; Kumar, Vikash ; Zaremba, Wojciech</creatorcontrib><description>The purpose of this technical report is two-fold. First of all, it introduces
a suite of challenging continuous control tasks (integrated with OpenAI Gym)
based on currently existing robotics hardware. The tasks include pushing,
sliding and pick & place with a Fetch robotic arm as well as in-hand object
manipulation with a Shadow Dexterous Hand. All tasks have sparse binary rewards
and follow a Multi-Goal Reinforcement Learning (RL) framework in which an agent
is told what to do using an additional input.
The second part of the paper presents a set of concrete research ideas for
improving RL algorithms, most of which are related to Multi-Goal RL and
Hindsight Experience Replay.</description><identifier>DOI: 10.48550/arxiv.1802.09464</identifier><language>eng</language><subject>Computer Science - Artificial Intelligence ; Computer Science - Learning ; Computer Science - Robotics</subject><creationdate>2018-02</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-a1154-3215e109ff18781f686f3c199756865e8bfa639e36f071ddef43082bc7173f4c3</citedby></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/1802.09464$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.1802.09464$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Plappert, Matthias</creatorcontrib><creatorcontrib>Andrychowicz, Marcin</creatorcontrib><creatorcontrib>Ray, Alex</creatorcontrib><creatorcontrib>McGrew, Bob</creatorcontrib><creatorcontrib>Baker, Bowen</creatorcontrib><creatorcontrib>Powell, Glenn</creatorcontrib><creatorcontrib>Schneider, Jonas</creatorcontrib><creatorcontrib>Tobin, Josh</creatorcontrib><creatorcontrib>Chociej, Maciek</creatorcontrib><creatorcontrib>Welinder, Peter</creatorcontrib><creatorcontrib>Kumar, Vikash</creatorcontrib><creatorcontrib>Zaremba, Wojciech</creatorcontrib><title>Multi-Goal Reinforcement Learning: Challenging Robotics Environments and Request for Research</title><description>The purpose of this technical report is two-fold. First of all, it introduces
a suite of challenging continuous control tasks (integrated with OpenAI Gym)
based on currently existing robotics hardware. The tasks include pushing,
sliding and pick & place with a Fetch robotic arm as well as in-hand object
manipulation with a Shadow Dexterous Hand. All tasks have sparse binary rewards
and follow a Multi-Goal Reinforcement Learning (RL) framework in which an agent
is told what to do using an additional input.
The second part of the paper presents a set of concrete research ideas for
improving RL algorithms, most of which are related to Multi-Goal RL and
Hindsight Experience Replay.</description><subject>Computer Science - Artificial Intelligence</subject><subject>Computer Science - Learning</subject><subject>Computer Science - Robotics</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2018</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotj0FOwzAURL1hgQoHYFVfIME_thOHHYpKQQpCqugSRY7z3VpKbXDSCm6PW1jNjDQz0iPkDlgulJTsXsdvd8pBsSJntSjFNfl4PY6zy9ZBj3SDztsQDR7Qz7RFHb3zuwfa7PU4ot-lQDehD7MzE135k4vBn6sT1X5I668jTjNND8lPaW32N-TK6nHC239dkO3T6r15ztq39Uvz2GYaQIqMFyARWG0tqEqBLVVpuYG6rmSyElVvdclr5KVlFQwDWsGZKnpTQcWtMHxBln-_F8DuM7qDjj_dGbS7gPJfGpdOSQ</recordid><startdate>20180226</startdate><enddate>20180226</enddate><creator>Plappert, Matthias</creator><creator>Andrychowicz, Marcin</creator><creator>Ray, Alex</creator><creator>McGrew, Bob</creator><creator>Baker, Bowen</creator><creator>Powell, Glenn</creator><creator>Schneider, Jonas</creator><creator>Tobin, Josh</creator><creator>Chociej, Maciek</creator><creator>Welinder, Peter</creator><creator>Kumar, Vikash</creator><creator>Zaremba, Wojciech</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20180226</creationdate><title>Multi-Goal Reinforcement Learning: Challenging Robotics Environments and Request for Research</title><author>Plappert, Matthias ; Andrychowicz, Marcin ; Ray, Alex ; McGrew, Bob ; Baker, Bowen ; Powell, Glenn ; Schneider, Jonas ; Tobin, Josh ; Chociej, Maciek ; Welinder, Peter ; Kumar, Vikash ; Zaremba, Wojciech</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a1154-3215e109ff18781f686f3c199756865e8bfa639e36f071ddef43082bc7173f4c3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2018</creationdate><topic>Computer Science - Artificial Intelligence</topic><topic>Computer Science - Learning</topic><topic>Computer Science - Robotics</topic><toplevel>online_resources</toplevel><creatorcontrib>Plappert, Matthias</creatorcontrib><creatorcontrib>Andrychowicz, Marcin</creatorcontrib><creatorcontrib>Ray, Alex</creatorcontrib><creatorcontrib>McGrew, Bob</creatorcontrib><creatorcontrib>Baker, Bowen</creatorcontrib><creatorcontrib>Powell, Glenn</creatorcontrib><creatorcontrib>Schneider, Jonas</creatorcontrib><creatorcontrib>Tobin, Josh</creatorcontrib><creatorcontrib>Chociej, Maciek</creatorcontrib><creatorcontrib>Welinder, Peter</creatorcontrib><creatorcontrib>Kumar, Vikash</creatorcontrib><creatorcontrib>Zaremba, Wojciech</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Plappert, Matthias</au><au>Andrychowicz, Marcin</au><au>Ray, Alex</au><au>McGrew, Bob</au><au>Baker, Bowen</au><au>Powell, Glenn</au><au>Schneider, Jonas</au><au>Tobin, Josh</au><au>Chociej, Maciek</au><au>Welinder, Peter</au><au>Kumar, Vikash</au><au>Zaremba, Wojciech</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Multi-Goal Reinforcement Learning: Challenging Robotics Environments and Request for Research</atitle><date>2018-02-26</date><risdate>2018</risdate><abstract>The purpose of this technical report is two-fold. First of all, it introduces
a suite of challenging continuous control tasks (integrated with OpenAI Gym)
based on currently existing robotics hardware. The tasks include pushing,
sliding and pick & place with a Fetch robotic arm as well as in-hand object
manipulation with a Shadow Dexterous Hand. All tasks have sparse binary rewards
and follow a Multi-Goal Reinforcement Learning (RL) framework in which an agent
is told what to do using an additional input.
The second part of the paper presents a set of concrete research ideas for
improving RL algorithms, most of which are related to Multi-Goal RL and
Hindsight Experience Replay.</abstract><doi>10.48550/arxiv.1802.09464</doi><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | DOI: 10.48550/arxiv.1802.09464 |
| ispartof | |
| issn | |
| language | eng |
| recordid | cdi_arxiv_primary_1802_09464 |
| source | arXiv.org |
| subjects | Computer Science - Artificial Intelligence Computer Science - Learning Computer Science - Robotics |
| title | Multi-Goal Reinforcement Learning: Challenging Robotics Environments and Request for Research |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-10T09%3A57%3A05IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Multi-Goal%20Reinforcement%20Learning:%20Challenging%20Robotics%20Environments%20and%20Request%20for%20Research&rft.au=Plappert,%20Matthias&rft.date=2018-02-26&rft_id=info:doi/10.48550/arxiv.1802.09464&rft_dat=%3Carxiv_GOX%3E1802_09464%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true |