Hyper-Parameter Selection on Convolutional Dictionary Learning Through Local ℓ0, ∞ Norm.
Gustavo SilvaJorge QuesadaPaul RodríguezPublished in: EUSIPCO (2019)
Keyphrases
- dictionary learning
- sparse coding
- sparse representation
- hyperparameters
- natural images
- image classification
- unsupervised learning
- linear combination
- generative model
- model selection
- random sampling
- parameter settings
- cross validation
- image patches
- visual words
- maximum a posteriori
- image representation
- closed form
- em algorithm
- maximum likelihood
- signal processing
- bayesian framework
- bayesian inference
- support vector
- learning algorithm
- semi supervised
- image features
- high dimensional
- feature extraction