A Majorize-Minimize Memory Gradient method for complex-valued inverse problems.
Anisia FlorescuÉmilie ChouzenouxJean-Christophe PesquetPhilippe CiuciuSilviu CiochinaPublished in: Signal Process. (2014)
Keyphrases
- complex valued
- gradient method
- inverse problems
- optimization methods
- real valued
- global optimization
- image reconstruction
- recurrent neural networks
- convex optimization
- convergence rate
- optimization problems
- optimization method
- linear equations
- step size
- simulated annealing
- multi valued
- activation function
- negative matrix factorization
- partial differential equations
- neural network
- early vision
- feature vectors
- computer vision
- multiscale
- pairwise
- multi objective
- metaheuristic