Roominoes: Generating Novel 3D Floor Plans From Existing 3D Rooms

Simple item page
dc.contributor.authorWang, Kaien_US
dc.contributor.authorXu, Xianghaoen_US
dc.contributor.authorLei, Leonen_US
dc.contributor.authorLing, Selenaen_US
dc.contributor.authorLindsay, Natalieen_US
dc.contributor.authorChang, Angel Xuanen_US
dc.contributor.authorSavva, Manolisen_US
dc.contributor.authorRitchie, Danielen_US
dc.contributor.editorDigne, Julie and Crane, Keenanen_US
dc.date.accessioned2021-07-10T07:46:15Z
dc.date.available2021-07-10T07:46:15Z
dc.date.issued2021
dc.description.abstractRealistic 3D indoor scene datasets have enabled significant recent progress in computer vision, scene understanding, autonomous navigation, and 3D reconstruction. But the scale, diversity, and customizability of existing datasets is limited, and it is time-consuming and expensive to scan and annotate more. Fortunately, combinatorics is on our side: there are enough individual rooms in existing 3D scene datasets, if there was but a way to recombine them into new layouts. In this paper, we propose the task of generating novel 3D floor plans from existing 3D rooms. We identify three sub-tasks of this problem: generation of 2D layout, retrieval of compatible 3D rooms, and deformation of 3D rooms to fit the layout. We then discuss different strategies for solving the problem, and design two representative pipelines: one uses available 2D floor plans to guide selection and deformation of 3D rooms; the other learns to retrieve a set of compatible 3D rooms and combine them into novel layouts. We design a set of metrics that evaluate the generated results with respect to each of the three subtasks and show that different methods trade off performance on these subtasks. Finally, we survey downstream tasks that benefit from generated 3D scenes and discuss strategies in selecting the methods most appropriate for the demands of these tasks.en_US
dc.description.number5
dc.description.sectionheadersShape Synthesis and Editing
dc.description.seriesinformationComputer Graphics Forum
dc.description.volume40
dc.identifier.doi10.1111/cgf.14357
dc.identifier.issn1467-8659
dc.identifier.pages57-69
dc.identifier.urihttps://doi.org/10.1111/cgf.14357
dc.identifier.urihttps://diglib.eg.org:443/handle/10.1111/cgf14357
dc.publisherThe Eurographics Association and John Wiley & Sons Ltd.en_US
dc.subjectComputing methodologies
dc.subjectComputer graphics
dc.subjectNeural networks
dc.subjectProbabilistic reasoning
dc.titleRoominoes: Generating Novel 3D Floor Plans From Existing 3D Roomsen_US
Files
Original bundle
Now showing 1 - 5 of 6
Thumbnail Image
Name:
v40i5pp057-069.pdf
Size:
8.68 MB
Format:
Adobe Portable Document Format
Download
No Thumbnail Available
Name:
ddppo_match_2d.mp4
Size:
74.64 KB
Format:
Unknown data format
Download
No Thumbnail Available
Name:
ddppo_mp3d.mp4
Size:
87.68 KB
Format:
Unknown data format
Download
No Thumbnail Available
Name:
ddppo_portal_stitching.mp4
Size:
83.91 KB
Format:
Unknown data format
Download
No Thumbnail Available
Name:
first_person_walkthrough.mp4
Size:
45.32 MB
Format:
Unknown data format
Download
Collections
40-Issue 5