Learning to Deblur using Light Field Generated and Real Defocus Images.
Lingyan RuanBin ChenJizhou LiMiu-Ling LamPublished in: CVPR (2022)
Keyphrases
- light field
- three dimensional
- acquired images
- object recognition
- active learning
- captured images
- synthetic and real images
- face recognition across pose
- learning algorithm
- image based rendering
- image formation
- input image
- geometric constraints
- multiple images
- imaging systems
- multi view
- optical flow
- epipolar plane
- viewpoint