AR-NeRF: Unsupervised Learning of Depth and Defocus Effects from Natural Images with Aperture Rendering Neural Radiance Fields.
Takuhiro KanekoPublished in: CVPR (2022)
Keyphrases
- natural images
- global illumination
- unsupervised learning
- depth from defocus
- object recognition
- computer graphics
- natural scenes
- higher order
- shape recovery
- defocus blur
- light field
- input image
- denoising
- multiscale
- sparse coding
- depth estimation
- depth map
- primary visual cortex
- spatial domain
- image patches
- visual cortex
- supervised learning
- receptive fields
- reflectance properties
- augmented reality
- neural network
- computer vision
- power spectrum
- image statistics
- statistics of natural images
- prior models
- biologically inspired
- photorealistic
- semi supervised
- deep learning
- natural image statistics
- dimensionality reduction
- natural image patches
- blurred images
- image processing