Cold Diffusion on the Replay Buffer: Learning to Plan from Known Good States
Learning from demonstrations (LfD) has successfully trained robots to exhibit remarkable generalization capabilities. However, many powerful imitation techniques do not prioritize the feasibility of the robot behaviors they generate. In this work, we explore the feasibility of plans produced by LfD....
Gespeichert in:
Veröffentlicht in: | arXiv.org 2023-10 |
---|---|
Hauptverfasser: | , , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | |
---|---|
container_issue | |
container_start_page | |
container_title | arXiv.org |
container_volume | |
creator | Wang, Zidan Oba, Takeru Yoneda, Takuma Shen, Rui Walter, Matthew Stadie, Bradly C |
description | Learning from demonstrations (LfD) has successfully trained robots to exhibit remarkable generalization capabilities. However, many powerful imitation techniques do not prioritize the feasibility of the robot behaviors they generate. In this work, we explore the feasibility of plans produced by LfD. As in prior work, we employ a temporal diffusion model with fixed start and goal states to facilitate imitation through in-painting. Unlike previous studies, we apply cold diffusion to ensure the optimization process is directed through the agent's replay buffer of previously visited states. This routing approach increases the likelihood that the final trajectories will predominantly occupy the feasible region of the robot's state space. We test this method in simulated robotic environments with obstacles and observe a significant improvement in the agent's ability to avoid these obstacles during planning. |
format | Article |
fullrecord | <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2881056397</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2881056397</sourcerecordid><originalsourceid>FETCH-proquest_journals_28810563973</originalsourceid><addsrcrecordid>eNqNiksKwjAUAIMgWLR3eOC6kCb2o0vrD3Qh6r4EmmhLzKv5IN7eLjyAMDCLmRGJGOdpUi4Ym5DYuY5SyvKCZRmPyKlC3cCmVSq4Fg0M-IeEi-y1-MA6KCXtCk5SWNOaO3iEsxYGlMUnHA2-DewRG7h64aWbkbES2sn45ymZ77a36pD0Fl9BOl93GKwZUs3KMqVZzpcF_-_6ArEUPFc</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2881056397</pqid></control><display><type>article</type><title>Cold Diffusion on the Replay Buffer: Learning to Plan from Known Good States</title><source>Free E- Journals</source><creator>Wang, Zidan ; Oba, Takeru ; Yoneda, Takuma ; Shen, Rui ; Walter, Matthew ; Stadie, Bradly C</creator><creatorcontrib>Wang, Zidan ; Oba, Takeru ; Yoneda, Takuma ; Shen, Rui ; Walter, Matthew ; Stadie, Bradly C</creatorcontrib><description>Learning from demonstrations (LfD) has successfully trained robots to exhibit remarkable generalization capabilities. However, many powerful imitation techniques do not prioritize the feasibility of the robot behaviors they generate. In this work, we explore the feasibility of plans produced by LfD. As in prior work, we employ a temporal diffusion model with fixed start and goal states to facilitate imitation through in-painting. Unlike previous studies, we apply cold diffusion to ensure the optimization process is directed through the agent's replay buffer of previously visited states. This routing approach increases the likelihood that the final trajectories will predominantly occupy the feasible region of the robot's state space. We test this method in simulated robotic environments with obstacles and observe a significant improvement in the agent's ability to avoid these obstacles during planning.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Barriers ; Buffers ; Feasibility ; Learning ; Optimization ; Robots</subject><ispartof>arXiv.org, 2023-10</ispartof><rights>2023. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>781,785</link.rule.ids></links><search><creatorcontrib>Wang, Zidan</creatorcontrib><creatorcontrib>Oba, Takeru</creatorcontrib><creatorcontrib>Yoneda, Takuma</creatorcontrib><creatorcontrib>Shen, Rui</creatorcontrib><creatorcontrib>Walter, Matthew</creatorcontrib><creatorcontrib>Stadie, Bradly C</creatorcontrib><title>Cold Diffusion on the Replay Buffer: Learning to Plan from Known Good States</title><title>arXiv.org</title><description>Learning from demonstrations (LfD) has successfully trained robots to exhibit remarkable generalization capabilities. However, many powerful imitation techniques do not prioritize the feasibility of the robot behaviors they generate. In this work, we explore the feasibility of plans produced by LfD. As in prior work, we employ a temporal diffusion model with fixed start and goal states to facilitate imitation through in-painting. Unlike previous studies, we apply cold diffusion to ensure the optimization process is directed through the agent's replay buffer of previously visited states. This routing approach increases the likelihood that the final trajectories will predominantly occupy the feasible region of the robot's state space. We test this method in simulated robotic environments with obstacles and observe a significant improvement in the agent's ability to avoid these obstacles during planning.</description><subject>Barriers</subject><subject>Buffers</subject><subject>Feasibility</subject><subject>Learning</subject><subject>Optimization</subject><subject>Robots</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><recordid>eNqNiksKwjAUAIMgWLR3eOC6kCb2o0vrD3Qh6r4EmmhLzKv5IN7eLjyAMDCLmRGJGOdpUi4Ym5DYuY5SyvKCZRmPyKlC3cCmVSq4Fg0M-IeEi-y1-MA6KCXtCk5SWNOaO3iEsxYGlMUnHA2-DewRG7h64aWbkbES2sn45ymZ77a36pD0Fl9BOl93GKwZUs3KMqVZzpcF_-_6ArEUPFc</recordid><startdate>20231021</startdate><enddate>20231021</enddate><creator>Wang, Zidan</creator><creator>Oba, Takeru</creator><creator>Yoneda, Takuma</creator><creator>Shen, Rui</creator><creator>Walter, Matthew</creator><creator>Stadie, Bradly C</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20231021</creationdate><title>Cold Diffusion on the Replay Buffer: Learning to Plan from Known Good States</title><author>Wang, Zidan ; Oba, Takeru ; Yoneda, Takuma ; Shen, Rui ; Walter, Matthew ; Stadie, Bradly C</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_28810563973</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Barriers</topic><topic>Buffers</topic><topic>Feasibility</topic><topic>Learning</topic><topic>Optimization</topic><topic>Robots</topic><toplevel>online_resources</toplevel><creatorcontrib>Wang, Zidan</creatorcontrib><creatorcontrib>Oba, Takeru</creatorcontrib><creatorcontrib>Yoneda, Takuma</creatorcontrib><creatorcontrib>Shen, Rui</creatorcontrib><creatorcontrib>Walter, Matthew</creatorcontrib><creatorcontrib>Stadie, Bradly C</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science & Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Access via ProQuest (Open Access)</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Wang, Zidan</au><au>Oba, Takeru</au><au>Yoneda, Takuma</au><au>Shen, Rui</au><au>Walter, Matthew</au><au>Stadie, Bradly C</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>Cold Diffusion on the Replay Buffer: Learning to Plan from Known Good States</atitle><jtitle>arXiv.org</jtitle><date>2023-10-21</date><risdate>2023</risdate><eissn>2331-8422</eissn><abstract>Learning from demonstrations (LfD) has successfully trained robots to exhibit remarkable generalization capabilities. However, many powerful imitation techniques do not prioritize the feasibility of the robot behaviors they generate. In this work, we explore the feasibility of plans produced by LfD. As in prior work, we employ a temporal diffusion model with fixed start and goal states to facilitate imitation through in-painting. Unlike previous studies, we apply cold diffusion to ensure the optimization process is directed through the agent's replay buffer of previously visited states. This routing approach increases the likelihood that the final trajectories will predominantly occupy the feasible region of the robot's state space. We test this method in simulated robotic environments with obstacles and observe a significant improvement in the agent's ability to avoid these obstacles during planning.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record> |
fulltext | fulltext |
identifier | EISSN: 2331-8422 |
ispartof | arXiv.org, 2023-10 |
issn | 2331-8422 |
language | eng |
recordid | cdi_proquest_journals_2881056397 |
source | Free E- Journals |
subjects | Barriers Buffers Feasibility Learning Optimization Robots |
title | Cold Diffusion on the Replay Buffer: Learning to Plan from Known Good States |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-13T05%3A45%3A40IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=Cold%20Diffusion%20on%20the%20Replay%20Buffer:%20Learning%20to%20Plan%20from%20Known%20Good%20States&rft.jtitle=arXiv.org&rft.au=Wang,%20Zidan&rft.date=2023-10-21&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2881056397%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2881056397&rft_id=info:pmid/&rfr_iscdi=true |