Provably Convergent Data-Driven Convex-Nonconvex Regularization.
Zakhar ShumaylovJeremy BuddSubhadip MukherjeeCarola-Bibiane SchönliebPublished in: CoRR (2023)
Keyphrases
- data driven
- provably convergent
- convex optimization
- data fidelity term
- shape from shading
- total variation
- quadratic optimization problems
- convex formulation
- image restoration and reconstruction
- convex functions
- risk minimization
- total variation regularization
- minimization problems
- norm minimization
- regularization term
- alternating direction method of multipliers
- augmented lagrangian method
- convex relaxation
- mumford shah functional
- penalty functions
- augmented lagrangian
- objective function
- convex optimization problems
- denoising
- convex sets
- optimization problems
- strictly convex
- primal dual
- regularization parameter
- bregman divergences
- globally convergent
- rank minimization
- stationary points
- interior point methods
- convex hull
- low rank
- nonlinear programming
- image segmentation