Sammanfattning

Reconstruction of indoor surfaces with limited texture information or with repeated textures, a situation common in walls and ceilings, may be difficult with a monocular Structure from Motion system. We propose a Semantic Room Wireframe Detection task to predict a Semantic Wireframe from a single perspective image. Such predictions may be used with shape priors to estimate the Room Layout and aid reconstruction. To train and test the proposed algorithm we create a new set of annotations from the simulated Structured3D dataset. We show qualitatively that the SRW-Net handles complex room geometries better than previous Room Layout Estimation algorithms while quantitatively out-performing the baseline in non-semantic Wireframe Detection.
Originalspråkengelska
Titel på värdpublikation26th International Conference on Pattern Recognition, 2022
FörlagIEEE - Institute of Electrical and Electronics Engineers Inc.
Sidor1886-1893
ISBN (elektroniskt)978-1-6654-9062-7
ISBN (tryckt)978-1-6654-9063-4
DOI
StatusPublished - 2022 aug. 25
Evenemang26TH International Conference on Pattern Recognition, 2022 - Montreal, Kanada
Varaktighet: 2022 aug. 212022 aug. 25

Konferens

Konferens26TH International Conference on Pattern Recognition, 2022
Förkortad titelICPR 2022
Land/TerritoriumKanada
OrtMontreal
Period2022/08/212022/08/25

Ämnesklassifikation (UKÄ)

  • Datorseende och robotik (autonoma system)

Fingeravtryck

Utforska forskningsämnen för ”Semantic Room Wireframe Detection from a Single View”. Tillsammans bildar de ett unikt fingeravtryck.

Citera det här