PBNS: Physically Based Neural Simulator for Unsupervised Garment Pose Space Deformation

We present a methodology to automatically obtain Pose Space Deformation (PSD) basis for rigged garments through deep learning. Classical approaches rely on Physically Based Simulations (PBS) to animate clothes. These are general solutions that, given a sufficiently fine-grained discretization of spa...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Bertiche, Hugo, Madadi, Meysam, Escalera, Sergio
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 Bertiche, Hugo
Madadi, Meysam
Escalera, Sergio
description We present a methodology to automatically obtain Pose Space Deformation (PSD) basis for rigged garments through deep learning. Classical approaches rely on Physically Based Simulations (PBS) to animate clothes. These are general solutions that, given a sufficiently fine-grained discretization of space and time, can achieve highly realistic results. However, they are computationally expensive and any scene modification prompts the need of re-simulation. Linear Blend Skinning (LBS) with PSD offers a lightweight alternative to PBS, though, it needs huge volumes of data to learn proper PSD. We propose using deep learning, formulated as an implicit PBS, to unsupervisedly learn realistic cloth Pose Space Deformations in a constrained scenario: dressed humans. Furthermore, we show it is possible to train these models in an amount of time comparable to a PBS of a few sequences. To the best of our knowledge, we are the first to propose a neural simulator for cloth. While deep-based approaches in the domain are becoming a trend, these are data-hungry models. Moreover, authors often propose complex formulations to better learn wrinkles from PBS data. Supervised learning leads to physically inconsistent predictions that require collision solving to be used. Also, dependency on PBS data limits the scalability of these solutions, while their formulation hinders its applicability and compatibility. By proposing an unsupervised methodology to learn PSD for LBS models (3D animation standard), we overcome both of these drawbacks. Results obtained show cloth-consistency in the animated garments and meaningful pose-dependant folds and wrinkles. Our solution is extremely efficient, handles multiple layers of cloth, allows unsupervised outfit resizing and can be easily applied to any custom 3D avatar.
doi_str_mv 10.48550/arxiv.2012.11310
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2012_11310</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2012_11310</sourcerecordid><originalsourceid>FETCH-LOGICAL-a670-5f1c50b99c4119bb2c76c8a483ff97c2b573668b8a80638d2fe0ea61b96e67b73</originalsourceid><addsrcrecordid>eNotj71OwzAURr0woMIDMOEXSLDjxHa60QIFqSqRUtQxunavhaX8yUkq8va0heHTtxwd6RDywFmc6ixjTxB-_ClOGE9izgVnt-RQrHblkhbf8-At1PVMVzDgke5wClDT0jdTDWMXqDvvqx2mHsPJX4gNhAbbkRbdgLTswSJ9wTPVwOi79o7cOKgHvP__Bdm_ve7X79H2c_Oxft5GIBWLMsdtxkye25Tz3JjEKmk1pFo4lyubmEwJKbXRoJkU-pg4ZAiSm1yiVEaJBXn8017Lqj74BsJcXQqra6H4BT6kS48</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>PBNS: Physically Based Neural Simulator for Unsupervised Garment Pose Space Deformation</title><source>arXiv.org</source><creator>Bertiche, Hugo ; Madadi, Meysam ; Escalera, Sergio</creator><creatorcontrib>Bertiche, Hugo ; Madadi, Meysam ; Escalera, Sergio</creatorcontrib><description>We present a methodology to automatically obtain Pose Space Deformation (PSD) basis for rigged garments through deep learning. Classical approaches rely on Physically Based Simulations (PBS) to animate clothes. These are general solutions that, given a sufficiently fine-grained discretization of space and time, can achieve highly realistic results. However, they are computationally expensive and any scene modification prompts the need of re-simulation. Linear Blend Skinning (LBS) with PSD offers a lightweight alternative to PBS, though, it needs huge volumes of data to learn proper PSD. We propose using deep learning, formulated as an implicit PBS, to unsupervisedly learn realistic cloth Pose Space Deformations in a constrained scenario: dressed humans. Furthermore, we show it is possible to train these models in an amount of time comparable to a PBS of a few sequences. To the best of our knowledge, we are the first to propose a neural simulator for cloth. While deep-based approaches in the domain are becoming a trend, these are data-hungry models. Moreover, authors often propose complex formulations to better learn wrinkles from PBS data. Supervised learning leads to physically inconsistent predictions that require collision solving to be used. Also, dependency on PBS data limits the scalability of these solutions, while their formulation hinders its applicability and compatibility. By proposing an unsupervised methodology to learn PSD for LBS models (3D animation standard), we overcome both of these drawbacks. Results obtained show cloth-consistency in the animated garments and meaningful pose-dependant folds and wrinkles. Our solution is extremely efficient, handles multiple layers of cloth, allows unsupervised outfit resizing and can be easily applied to any custom 3D avatar.</description><identifier>DOI: 10.48550/arxiv.2012.11310</identifier><language>eng</language><subject>Computer Science - Computer Vision and Pattern Recognition ; Computer Science - Graphics</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/2012.11310$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2012.11310$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Bertiche, Hugo</creatorcontrib><creatorcontrib>Madadi, Meysam</creatorcontrib><creatorcontrib>Escalera, Sergio</creatorcontrib><title>PBNS: Physically Based Neural Simulator for Unsupervised Garment Pose Space Deformation</title><description>We present a methodology to automatically obtain Pose Space Deformation (PSD) basis for rigged garments through deep learning. Classical approaches rely on Physically Based Simulations (PBS) to animate clothes. These are general solutions that, given a sufficiently fine-grained discretization of space and time, can achieve highly realistic results. However, they are computationally expensive and any scene modification prompts the need of re-simulation. Linear Blend Skinning (LBS) with PSD offers a lightweight alternative to PBS, though, it needs huge volumes of data to learn proper PSD. We propose using deep learning, formulated as an implicit PBS, to unsupervisedly learn realistic cloth Pose Space Deformations in a constrained scenario: dressed humans. Furthermore, we show it is possible to train these models in an amount of time comparable to a PBS of a few sequences. To the best of our knowledge, we are the first to propose a neural simulator for cloth. While deep-based approaches in the domain are becoming a trend, these are data-hungry models. Moreover, authors often propose complex formulations to better learn wrinkles from PBS data. Supervised learning leads to physically inconsistent predictions that require collision solving to be used. Also, dependency on PBS data limits the scalability of these solutions, while their formulation hinders its applicability and compatibility. By proposing an unsupervised methodology to learn PSD for LBS models (3D animation standard), we overcome both of these drawbacks. Results obtained show cloth-consistency in the animated garments and meaningful pose-dependant folds and wrinkles. Our solution is extremely efficient, handles multiple layers of cloth, allows unsupervised outfit resizing and can be easily applied to any custom 3D avatar.</description><subject>Computer Science - Computer Vision and Pattern Recognition</subject><subject>Computer Science - Graphics</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2020</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotj71OwzAURr0woMIDMOEXSLDjxHa60QIFqSqRUtQxunavhaX8yUkq8va0heHTtxwd6RDywFmc6ixjTxB-_ClOGE9izgVnt-RQrHblkhbf8-At1PVMVzDgke5wClDT0jdTDWMXqDvvqx2mHsPJX4gNhAbbkRbdgLTswSJ9wTPVwOi79o7cOKgHvP__Bdm_ve7X79H2c_Oxft5GIBWLMsdtxkye25Tz3JjEKmk1pFo4lyubmEwJKbXRoJkU-pg4ZAiSm1yiVEaJBXn8017Lqj74BsJcXQqra6H4BT6kS48</recordid><startdate>20201221</startdate><enddate>20201221</enddate><creator>Bertiche, Hugo</creator><creator>Madadi, Meysam</creator><creator>Escalera, Sergio</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20201221</creationdate><title>PBNS: Physically Based Neural Simulator for Unsupervised Garment Pose Space Deformation</title><author>Bertiche, Hugo ; Madadi, Meysam ; Escalera, Sergio</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a670-5f1c50b99c4119bb2c76c8a483ff97c2b573668b8a80638d2fe0ea61b96e67b73</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2020</creationdate><topic>Computer Science - Computer Vision and Pattern Recognition</topic><topic>Computer Science - Graphics</topic><toplevel>online_resources</toplevel><creatorcontrib>Bertiche, Hugo</creatorcontrib><creatorcontrib>Madadi, Meysam</creatorcontrib><creatorcontrib>Escalera, Sergio</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Bertiche, Hugo</au><au>Madadi, Meysam</au><au>Escalera, Sergio</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>PBNS: Physically Based Neural Simulator for Unsupervised Garment Pose Space Deformation</atitle><date>2020-12-21</date><risdate>2020</risdate><abstract>We present a methodology to automatically obtain Pose Space Deformation (PSD) basis for rigged garments through deep learning. Classical approaches rely on Physically Based Simulations (PBS) to animate clothes. These are general solutions that, given a sufficiently fine-grained discretization of space and time, can achieve highly realistic results. However, they are computationally expensive and any scene modification prompts the need of re-simulation. Linear Blend Skinning (LBS) with PSD offers a lightweight alternative to PBS, though, it needs huge volumes of data to learn proper PSD. We propose using deep learning, formulated as an implicit PBS, to unsupervisedly learn realistic cloth Pose Space Deformations in a constrained scenario: dressed humans. Furthermore, we show it is possible to train these models in an amount of time comparable to a PBS of a few sequences. To the best of our knowledge, we are the first to propose a neural simulator for cloth. While deep-based approaches in the domain are becoming a trend, these are data-hungry models. Moreover, authors often propose complex formulations to better learn wrinkles from PBS data. Supervised learning leads to physically inconsistent predictions that require collision solving to be used. Also, dependency on PBS data limits the scalability of these solutions, while their formulation hinders its applicability and compatibility. By proposing an unsupervised methodology to learn PSD for LBS models (3D animation standard), we overcome both of these drawbacks. Results obtained show cloth-consistency in the animated garments and meaningful pose-dependant folds and wrinkles. Our solution is extremely efficient, handles multiple layers of cloth, allows unsupervised outfit resizing and can be easily applied to any custom 3D avatar.</abstract><doi>10.48550/arxiv.2012.11310</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2012.11310
ispartof
issn
language eng
recordid cdi_arxiv_primary_2012_11310
source arXiv.org
subjects Computer Science - Computer Vision and Pattern Recognition
Computer Science - Graphics
title PBNS: Physically Based Neural Simulator for Unsupervised Garment Pose Space Deformation
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-19T07%3A30%3A46IST&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=PBNS:%20Physically%20Based%20Neural%20Simulator%20for%20Unsupervised%20Garment%20Pose%20Space%20Deformation&rft.au=Bertiche,%20Hugo&rft.date=2020-12-21&rft_id=info:doi/10.48550/arxiv.2012.11310&rft_dat=%3Carxiv_GOX%3E2012_11310%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