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