SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single Image.
Dejia XuYifan JiangPeihao WangZhiwen FanHumphrey ShiZhangyang WangPublished in: CoRR (2022)
Keyphrases
- single image
- complex scenes
- multiple images
- object surface
- shape recovery
- image pairs
- light source
- d scene
- shape from shading
- super resolution
- computer graphics
- photorealistic
- multiple objects
- indoor scenes
- partially occluded
- reflectance properties
- intrinsic images
- vanishing points
- structured light
- intrinsic image decomposition
- machine learning
- information from multiple images
- viewpoint
- computer vision