DisCoScene: Spatially Disentangled Generative Radiance Fields for Controllable 3D-aware Scene Synthesis.
Yinghao XuMenglei ChaiZifan ShiSida PengIvan SkorokhodovAliaksandr SiarohinCeyuan YangYujun ShenHsin-Ying LeeBolei ZhouSergey TulyakovPublished in: CVPR (2023)
Keyphrases
- global illumination
- d scene
- generative model
- dynamic scenes
- multiple images
- video sequences
- single image
- real scenes
- three dimensional
- camera response function
- image sequences
- ray tracing
- complex scenes
- scene classification
- photometric properties
- outdoor scenes
- spatial location
- discriminative learning
- input image
- spectral reflectance
- moving objects
- image based rendering
- lighting conditions