Problem-dependent attention and effort in neural networks with applications to image resolution and model selection.
Chris RohlfsPublished in: Image Vis. Comput. (2023)
Keyphrases
- model selection
- image resolution
- image quality
- field of view
- cross validation
- spatial resolution
- super resolution
- parameter estimation
- hyperparameters
- bayesian learning
- regression model
- sample size
- mixture model
- high resolution
- machine learning
- generalization error
- low resolution
- motion segmentation
- pattern recognition
- model selection criteria
- gaussian process
- statistical inference
- information criterion
- variable selection
- image reconstruction
- marginal likelihood
- feature selection
- parameter determination
- high quality
- automatic model selection
- multispectral
- spatial information
- maximum likelihood
- decision trees