BuilDiff: 3D Building Shape Generation using Single-Image Conditional Point Cloud Diffusion Models.

Yao Wei George Vosselman Michael Ying Yang

Published in: ICCV (Workshops) (2023)

Keyphrases

single image
point cloud
diffusion models
shape recovery
diffusion model
structure from motion
intrinsic images
multiple images
surface reconstruction
image pairs
object surface
information diffusion
light source
social networks
d scene
point sets
shape from shading
super resolution
object recognition
image sequences