Configurable 3D Scene Synthesis and 2D Image Rendering with Per-pixel Ground Truth Using Stochastic Grammars.
Chenfanfu JiangSiyuan QiYixin ZhuSiyuan HuangJenny LinLap-Fai YuDemetri TerzopoulosSong-Chun ZhuPublished in: Int. J. Comput. Vis. (2018)
Keyphrases
- d scene
- image rendering
- ground truth
- single image
- depth map
- image based rendering
- high dynamic range
- high quality
- scene understanding
- optical flow
- camera viewpoint
- camera parameters
- view synthesis
- coordinate frame
- depth information
- real time
- flow estimation
- image processing
- scene flow
- machine learning
- viewpoint
- multiscale
- face recognition