Relightify: Relightable 3D Faces from a Single Image via Diffusion Models.
Foivos Paraperas PapantoniouAlexandros LattasStylianos MoschoglouStefanos ZafeiriouPublished in: ICCV (2023)
Keyphrases
- single image
- diffusion models
- diffusion model
- information diffusion
- partially occluded
- social networks
- multiple images
- d scene
- light source
- shape recovery
- human faces
- super resolution
- d morphable model
- viral marketing
- influence maximization
- face recognition
- indoor scenes
- line drawings
- anisotropic diffusion
- shape from shading
- face images
- online social networks
- image sequences
- image processing