Wall grid structure for interior scene synthesis

Handle URI:
http://hdl.handle.net/10754/564028
Title:
Wall grid structure for interior scene synthesis
Authors:
Xu, Wenzhuo; Wang, Bin; Yan, Dongming ( 0000-0003-2209-2404 )
Abstract:
We present a system for automatically synthesizing a diverse set of semantically valid, and well-arranged 3D interior scenes for a given empty room shape. Unlike existing work on layout synthesis, that typically knows potentially needed 3D models and optimizes their location through cost functions, our technique performs the retrieval and placement of 3D models by discovering the relationships between the room space and the models' categories. This is enabled by a new analytical structure, called Wall Grid Structure, which jointly considers the categories and locations of 3D models. Our technique greatly reduces the amount of user intervention and provides users with suggestions and inspirations. We demonstrate the applicability of our approach on three types of scenarios: conference rooms, living rooms and bedrooms.
KAUST Department:
Visual Computing Center (VCC)
Publisher:
Elsevier BV
Journal:
Computers & Graphics
Issue Date:
Feb-2015
DOI:
10.1016/j.cag.2014.09.032
Type:
Article
ISSN:
00978493
Sponsors:
This work was supported by the National Basic Research Program of China (2010CB328001), the National Science Foundation of China (61373071, 61372168), and the KAUST Visual Computing Center.
Appears in Collections:
Articles; Visual Computing Center (VCC)

Full metadata record

DC FieldValue Language
dc.contributor.authorXu, Wenzhuoen
dc.contributor.authorWang, Binen
dc.contributor.authorYan, Dongmingen
dc.date.accessioned2015-08-03T12:29:00Zen
dc.date.available2015-08-03T12:29:00Zen
dc.date.issued2015-02en
dc.identifier.issn00978493en
dc.identifier.doi10.1016/j.cag.2014.09.032en
dc.identifier.urihttp://hdl.handle.net/10754/564028en
dc.description.abstractWe present a system for automatically synthesizing a diverse set of semantically valid, and well-arranged 3D interior scenes for a given empty room shape. Unlike existing work on layout synthesis, that typically knows potentially needed 3D models and optimizes their location through cost functions, our technique performs the retrieval and placement of 3D models by discovering the relationships between the room space and the models' categories. This is enabled by a new analytical structure, called Wall Grid Structure, which jointly considers the categories and locations of 3D models. Our technique greatly reduces the amount of user intervention and provides users with suggestions and inspirations. We demonstrate the applicability of our approach on three types of scenarios: conference rooms, living rooms and bedrooms.en
dc.description.sponsorshipThis work was supported by the National Basic Research Program of China (2010CB328001), the National Science Foundation of China (61373071, 61372168), and the KAUST Visual Computing Center.en
dc.publisherElsevier BVen
dc.subjectFront extractionen
dc.subjectInterior scene synthesisen
dc.subjectWall grid structureen
dc.titleWall grid structure for interior scene synthesisen
dc.typeArticleen
dc.contributor.departmentVisual Computing Center (VCC)en
dc.identifier.journalComputers & Graphicsen
dc.contributor.institutionSchool of Software, Tsinghua UniversityBeijing, Chinaen
dc.contributor.institutionTsinghua National Laboratory for Information Science and TechnologyBeijing, Chinaen
dc.contributor.institutionNLPR-LIAMA, Institute of Automation, Chinese Academy of SciencesBeijing, Chinaen
kaust.authorYan, Dongmingen
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.