Cost-to-Go Function Generating Networks for High Dimensional Motion Planning

This paper presents c2g-HOF networks which learn to generate cost-to-go functions for manipulator motion planning. The c2g-HOF architecture consists of a cost-to-go function over the configuration space represented as a neural network (c2g-network) as well as a Higher Order Function (HOF) network wh...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Huh, Jinwook, Isler, Volkan, Lee, Daniel D
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 Huh, Jinwook
Isler, Volkan
Lee, Daniel D
description This paper presents c2g-HOF networks which learn to generate cost-to-go functions for manipulator motion planning. The c2g-HOF architecture consists of a cost-to-go function over the configuration space represented as a neural network (c2g-network) as well as a Higher Order Function (HOF) network which outputs the weights of the c2g-network for a given input workspace. Both networks are trained end-to-end in a supervised fashion using costs computed from traditional motion planners. Once trained, c2g-HOF can generate a smooth and continuous cost-to-go function directly from workspace sensor inputs (represented as a point cloud in 3D or an image in 2D). At inference time, the weights of the c2g-network are computed very efficiently and near-optimal trajectories are generated by simply following the gradient of the cost-to-go function. We compare c2g-HOF with traditional planning algorithms for various robots and planning scenarios. The experimental results indicate that planning with c2g-HOF is significantly faster than other motion planning algorithms, resulting in orders of magnitude improvement when including collision checking. Furthermore, despite being trained from sparsely sampled trajectories in configuration space, c2g-HOF generalizes to generate smoother, and often lower cost, trajectories. We demonstrate cost-to-go based planning on a 7 DoF manipulator arm where motion planning in a complex workspace requires only 0.13 seconds for the entire trajectory.
doi_str_mv 10.48550/arxiv.2012.06023
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2012_06023</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2012_06023</sourcerecordid><originalsourceid>FETCH-LOGICAL-a673-dafeef490d411431b06f96757f188d3513777ae53a526c129494c98d277391e93</originalsourceid><addsrcrecordid>eNotz7tOwzAYhmEvDKhwAUz4Bhx8djyiQFOkcBi6R38Tu7VIbeSY091TAtO3PPqkF6ErRitZK0VvIH-Fj4pTxiuqKRfnqGvSXEhJpE14_R6HElLErYsuQwlxj59c-Uz5dcY-ZbwJ-wO-C0cX5xODCT-mxb9MEONJX6AzD9PsLv93hbbr-22zId1z-9DcdgS0EWQE75yXlo6SMSnYjmpvtVHGs7oehWLCGANOCVBcD4xbaeVg65EbIyxzVqzQ9d_tktO_5XCE_N3_ZvVLlvgBikFGzw</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Cost-to-Go Function Generating Networks for High Dimensional Motion Planning</title><source>arXiv.org</source><creator>Huh, Jinwook ; Isler, Volkan ; Lee, Daniel D</creator><creatorcontrib>Huh, Jinwook ; Isler, Volkan ; Lee, Daniel D</creatorcontrib><description>This paper presents c2g-HOF networks which learn to generate cost-to-go functions for manipulator motion planning. The c2g-HOF architecture consists of a cost-to-go function over the configuration space represented as a neural network (c2g-network) as well as a Higher Order Function (HOF) network which outputs the weights of the c2g-network for a given input workspace. Both networks are trained end-to-end in a supervised fashion using costs computed from traditional motion planners. Once trained, c2g-HOF can generate a smooth and continuous cost-to-go function directly from workspace sensor inputs (represented as a point cloud in 3D or an image in 2D). At inference time, the weights of the c2g-network are computed very efficiently and near-optimal trajectories are generated by simply following the gradient of the cost-to-go function. We compare c2g-HOF with traditional planning algorithms for various robots and planning scenarios. The experimental results indicate that planning with c2g-HOF is significantly faster than other motion planning algorithms, resulting in orders of magnitude improvement when including collision checking. Furthermore, despite being trained from sparsely sampled trajectories in configuration space, c2g-HOF generalizes to generate smoother, and often lower cost, trajectories. We demonstrate cost-to-go based planning on a 7 DoF manipulator arm where motion planning in a complex workspace requires only 0.13 seconds for the entire trajectory.</description><identifier>DOI: 10.48550/arxiv.2012.06023</identifier><language>eng</language><subject>Computer Science - Artificial Intelligence ; Computer Science - Learning ; Computer Science - Robotics</subject><creationdate>2020-12</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,780,885</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2012.06023$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2012.06023$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Huh, Jinwook</creatorcontrib><creatorcontrib>Isler, Volkan</creatorcontrib><creatorcontrib>Lee, Daniel D</creatorcontrib><title>Cost-to-Go Function Generating Networks for High Dimensional Motion Planning</title><description>This paper presents c2g-HOF networks which learn to generate cost-to-go functions for manipulator motion planning. The c2g-HOF architecture consists of a cost-to-go function over the configuration space represented as a neural network (c2g-network) as well as a Higher Order Function (HOF) network which outputs the weights of the c2g-network for a given input workspace. Both networks are trained end-to-end in a supervised fashion using costs computed from traditional motion planners. Once trained, c2g-HOF can generate a smooth and continuous cost-to-go function directly from workspace sensor inputs (represented as a point cloud in 3D or an image in 2D). At inference time, the weights of the c2g-network are computed very efficiently and near-optimal trajectories are generated by simply following the gradient of the cost-to-go function. We compare c2g-HOF with traditional planning algorithms for various robots and planning scenarios. The experimental results indicate that planning with c2g-HOF is significantly faster than other motion planning algorithms, resulting in orders of magnitude improvement when including collision checking. Furthermore, despite being trained from sparsely sampled trajectories in configuration space, c2g-HOF generalizes to generate smoother, and often lower cost, trajectories. We demonstrate cost-to-go based planning on a 7 DoF manipulator arm where motion planning in a complex workspace requires only 0.13 seconds for the entire trajectory.</description><subject>Computer Science - Artificial Intelligence</subject><subject>Computer Science - Learning</subject><subject>Computer Science - Robotics</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2020</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotz7tOwzAYhmEvDKhwAUz4Bhx8djyiQFOkcBi6R38Tu7VIbeSY091TAtO3PPqkF6ErRitZK0VvIH-Fj4pTxiuqKRfnqGvSXEhJpE14_R6HElLErYsuQwlxj59c-Uz5dcY-ZbwJ-wO-C0cX5xODCT-mxb9MEONJX6AzD9PsLv93hbbr-22zId1z-9DcdgS0EWQE75yXlo6SMSnYjmpvtVHGs7oehWLCGANOCVBcD4xbaeVg65EbIyxzVqzQ9d_tktO_5XCE_N3_ZvVLlvgBikFGzw</recordid><startdate>20201210</startdate><enddate>20201210</enddate><creator>Huh, Jinwook</creator><creator>Isler, Volkan</creator><creator>Lee, Daniel D</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20201210</creationdate><title>Cost-to-Go Function Generating Networks for High Dimensional Motion Planning</title><author>Huh, Jinwook ; Isler, Volkan ; Lee, Daniel D</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a673-dafeef490d411431b06f96757f188d3513777ae53a526c129494c98d277391e93</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2020</creationdate><topic>Computer Science - Artificial Intelligence</topic><topic>Computer Science - Learning</topic><topic>Computer Science - Robotics</topic><toplevel>online_resources</toplevel><creatorcontrib>Huh, Jinwook</creatorcontrib><creatorcontrib>Isler, Volkan</creatorcontrib><creatorcontrib>Lee, Daniel D</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Huh, Jinwook</au><au>Isler, Volkan</au><au>Lee, Daniel D</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Cost-to-Go Function Generating Networks for High Dimensional Motion Planning</atitle><date>2020-12-10</date><risdate>2020</risdate><abstract>This paper presents c2g-HOF networks which learn to generate cost-to-go functions for manipulator motion planning. The c2g-HOF architecture consists of a cost-to-go function over the configuration space represented as a neural network (c2g-network) as well as a Higher Order Function (HOF) network which outputs the weights of the c2g-network for a given input workspace. Both networks are trained end-to-end in a supervised fashion using costs computed from traditional motion planners. Once trained, c2g-HOF can generate a smooth and continuous cost-to-go function directly from workspace sensor inputs (represented as a point cloud in 3D or an image in 2D). At inference time, the weights of the c2g-network are computed very efficiently and near-optimal trajectories are generated by simply following the gradient of the cost-to-go function. We compare c2g-HOF with traditional planning algorithms for various robots and planning scenarios. The experimental results indicate that planning with c2g-HOF is significantly faster than other motion planning algorithms, resulting in orders of magnitude improvement when including collision checking. Furthermore, despite being trained from sparsely sampled trajectories in configuration space, c2g-HOF generalizes to generate smoother, and often lower cost, trajectories. We demonstrate cost-to-go based planning on a 7 DoF manipulator arm where motion planning in a complex workspace requires only 0.13 seconds for the entire trajectory.</abstract><doi>10.48550/arxiv.2012.06023</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2012.06023
ispartof
issn
language eng
recordid cdi_arxiv_primary_2012_06023
source arXiv.org
subjects Computer Science - Artificial Intelligence
Computer Science - Learning
Computer Science - Robotics
title Cost-to-Go Function Generating Networks for High Dimensional Motion Planning
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-25T22%3A29%3A22IST&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=Cost-to-Go%20Function%20Generating%20Networks%20for%20High%20Dimensional%20Motion%20Planning&rft.au=Huh,%20Jinwook&rft.date=2020-12-10&rft_id=info:doi/10.48550/arxiv.2012.06023&rft_dat=%3Carxiv_GOX%3E2012_06023%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