Depth Estimation Through a Generative Model of Light Field Synthesis.
Mehdi S. M. SajjadiRolf KöhlerBernhard SchölkopfMichael HirschPublished in: GCPR (2016)
Keyphrases
- generative model
- light field
- depth estimation
- super resolution
- multi view
- depth map
- low resolution
- dynamic scenes
- high resolution
- probabilistic model
- stereo vision
- semi supervised
- image formation
- image restoration
- stereo matching
- view synthesis
- em algorithm
- depth information
- motion estimation
- high quality
- multiple views
- single image
- multiple cameras
- d objects
- real scenes
- prior knowledge
- d scene
- motion blur
- expectation maximization
- stereo pair
- video camera
- image segmentation
- graphical models
- image registration
- dynamic programming
- similarity measure