Generalized SURE for optimal shrinkage of singular values in low-rank matrix denoising.
Jérémie BigotCharles DeledalleDelphine FéralPublished in: J. Mach. Learn. Res. (2017)
Keyphrases
- denoising
- singular values
- low rank matrix
- singular value decomposition
- low rank
- nuclear norm
- matrix decomposition
- image denoising
- total variation
- image processing
- matrix factorization
- natural images
- convex optimization
- wavelet domain
- linear combination
- matrix completion
- denoising algorithm
- sparse matrix
- semi supervised
- closed form
- rank minimization
- data sets
- similar patches
- least squares
- dynamic programming
- multiple views
- feature vectors
- object recognition
- multiscale