Batch-less stochastic gradient descent for compressive learning of deep regularization for image denoising.
Hui ShiYann TraonmilinJean-François AujolPublished in: CoRR (2023)
Keyphrases
- stochastic gradient descent
- image denoising
- denoising
- matrix factorization
- learning tasks
- noisy images
- learning algorithm
- natural images
- least squares
- loss function
- total variation
- step size
- regularization parameter
- online algorithms
- image reconstruction
- wavelet coefficients
- random forests
- multiple kernel learning
- feature space