MetaDIP: Accelerating Deep Image Prior with Meta Learning.
Kevin ZhangMingyang XieMaharshi GorYi-Ting ChenYvonne ZhouChristopher A. MetzlerPublished in: CoRR (2022)
Keyphrases
- meta learning
- image prior
- image restoration
- learning tasks
- edge preserving
- image denoising
- markov random field
- inductive learning
- reconstructed image
- model selection
- maximum a posteriori
- edge information
- blind deconvolution
- machine learning algorithms
- natural scenes
- image deblurring
- prior model
- decision trees
- machine learning
- additive noise
- base classifiers
- colour images
- image reconstruction
- feature selection
- noise reduction
- map estimation
- natural images
- image compression
- data sets
- potential functions
- knowledge acquisition
- bayesian networks
- bit rate
- learning algorithm
- maximum likelihood
- image quality
- feature set