PALMNUT: An Enhanced Proximal Alternating Linearized Minimization Algorithm With Application to Separate Regularization of Magnitude and Phase.
Yunsong LiuJustin P. HaldarPublished in: IEEE Trans. Computational Imaging (2021)
Keyphrases
- learning algorithm
- computational cost
- dynamic programming
- high accuracy
- worst case
- times faster
- theoretical analysis
- matching algorithm
- objective function
- cost function
- k means
- preprocessing
- experimental evaluation
- preprocessing phase
- kalman filter
- convergence rate
- image processing
- expectation maximization
- linear programming
- search space
- optimal solution
- segmentation algorithm
- detection algorithm
- np hard
- significant improvement
- regularization term
- learning phase
- similarity measure
- solution path