Look-Ahead Training with Learned Reflectance Loss for Single-Image SVBRDF Estimation.
Xilong ZhouNima Khademi KalantariPublished in: ACM Trans. Graph. (2022)
Keyphrases
- single image
- multiple images
- light source
- d scene
- image pairs
- intrinsic image decomposition
- shape recovery
- shape from shading
- training set
- intrinsic images
- partially occluded
- super resolution
- spherical harmonics
- information from multiple images
- scene geometry
- vanishing points
- estimation algorithm
- outdoor images
- structured light