Three variants of alternating direction method of multipliers without certain inner iterations and their application to image super-resolution via sparse representation
スポンサーリンク
概要
- 論文の詳細を見る
We propose three variants of Alternating Direction Method of Multipliers (ADMM) employing simplified updates under additional assumptions. ADMM iteratively solves the minimization of the sum of two nonsmooth convex functions. Each iteration of ADMM itself consists of solving a certain convex optimization problem which often requires the use of some iterative solver. These inner iterations cause slow convergence. Our proposed algorithms avoid certain inner iterations by employing simplified updates. Indeed, in several scenarios of image super-resolution via sparse representation, the proposed algorithms do not require matrix inversion which causes inner iterations in the standard ADMM. Numerical examples for the image super-resolution setting demonstrate that our proposed algorithms reduce CPU time to about 21-77 percent of the standard ADMM.
- 2012-03-01
著者
-
Yamada Isao
Departmen Of Communications And Integrated Systems Tokyo Institute Of Technology
-
Yamagishi Masao
Department Of Communications And Integrated Systems Tokyo Institute Of Technology
-
ONO Shunsuke
Department of Communications and Integrated Systems, Tokyo Institute of Technology
-
YAMAGISHI Masao
Department of Communications and Computer Engineering Tokyo Institute of Technology
-
ONO Shunsuke
Department of Communications and Computer Engineering Tokyo Institute of Technology
関連論文
- Pressure-Induced Transition of the Interlayer Exchange Interaction from Ferromagnetic to Antiferromagnetic Observed in the Two-Dimensional Ferromagnet (CH_3NH_3)_2 CuCl_4 through Magnetic Susceptibility and Neutron Scattering Measurements : Condensed Matt
- Observation of the Pressure-Induced Weak Ferromagnetic Moment in the Layered Antiferromagnetic Compound (C_2H_5NH_3)_2CuCl_4 by Magnetic Susceptibility and Neutron Scattering Measurements : Condensed Matter: Electronic Properties, etc.
- Temperature and Angular Dependence of Electron Paramagnetic Resonance Line in α'-NaV_2O_5
- Disappearance of the Weak Ferromagnetic Moment under High Pressure Observed in the Two-Dimensional Antiferromagnet (C_3H_7NH_3)_2CuCl_4 through Magnetic Susceptibility Measurements
- Electron Paramagnetic Resonance in the Singlet Dimer System (CH_3)_2CHNH_3CuBr_3 : Condensed Matter: Electronic Properties, etc.
- Field-Induced Magnetization in Ferromagnetic and Antiferromagnetic Alternating Heisenberg Chains Formed in (CH_3)_2CHNH_3CuCl_3 : Condensed Matter: Electronic Properties, etc.
- ^7Li NMR Study of the Triangular Lattice Antiferromagnet LiNiO_2
- ESR Study of the Triangular Lattice Antiferromagnets with S=1/2:NaTiO_2 and LiNiO_2
- ESR Study of the Triangular Lattice Heisenberg Antiferromagnets VX_2(X=Cl,Br and I)
- A Higher Order Generalization of an Alias-Free Discrete Time-Frequency Analysis(Special Section on Digital Signal Processing)
- A Note on Constrained Least Squares Design of M-D FIR Filter Based on Convex Projection Techniques(Special Section on Digital Signal Processing)
- CONSTRAINED LEAST SQUARES DESIGN OF M-D FIR FILTER BASED ON CONVEX PROJECTION TECHNIQUES
- Optical Birefringence Measurements in Ferromagnetic and Antiferromagnetic Alternating Heisenberg Chains (CH_3)_2CHNH_3CuCl_3
- Numerically stable algorithms for adaptive generalized minor subspace extraction (通信方式)
- Numerically stable algorithms for adaptive generalized minor subspace extraction (回路とシステム)
- Numerically stable algorithms for adaptive generalized minor subspace extraction (信号処理)
- Electron Spin Resonance in Ferromagnetic-Dominant Ferromagnetic-Antiferromagnetic Alternating Heisenberg Chains (4-BzpipdH) CuCl_3
- Electron Spin Resonance in Ferromagnetic-Dominant Ferromagnetic-Antiferromagnetic Alternating Heisenberg Chains (4-BzpipdH)CuCl_3
- Field-Induced Magnetic Long-Range Order in the Ferromagnetic-Antiferromagnetic Alternating Heisenberg Chain System (CH_3)_2CHNH_3CuCl_3 Observed by Specific Heat Measurements
- Observation of a Spin-Glass Transition in the S=1 Heisenberg Antiferromagnet Ni(C_2H_8N_2)_2NO_2(ClO_4)
- An Associative Memory Neural Network to Recall Nearest Pattern from Input
- Change of Magnetism in the Two-Dimensional Heisenberg Ferromagnet K_2CuF_4 Observed at High Pressures
- Magnetic Phase Transitions in the Two-Dimensional Random Mixture K_2Cu_xCo_F_4 with Computing Exchange Interactions
- Study of Magnetic Energy in One-Dimensional S=1/2 Heisenberg Antiferromagnet KCuF_3 by Optical Birefringence
- A Set-Theoretic Blind Image Deconvolution Based on Hybrid Steepest Descent Method (Special Section on Digital Signal Processing)
- Static Spin Correlations in a Diluted Two-Dimensional Ferromagnet K_2Cu_xZn_F_4 Including the Percolation Vicinity
- One-Dimensional Heisenberg Ferromagnetic Susceptibility Just at the Percolation Threshold for K_2Cu_xZn_F_4
- Critical Temperature of Randomly Diluted Two-Dimensional Heisenberg Ferromagnet, K_2Cu_xZn_F_4
- High Field Maganetization Process of Random Mixtures with Competing Exchange Interactions K_2Cu_xA_F_4(A=Mn and Co)
- DC Electroluminescence in Mn-Implanted ZnS Thin Films
- Neutron Scattering Study of Two-Dimensional Randomly Mixed Magnets with Competing Anisotropies K_2CuCo_xF_4
- Observation of Sidebands on the EPR Lines in Low-Dimensional Systems
- Observation of Pressure-Induced Structural Transition in K^3CuF^4 by Phonon-Raman Scattering Experiments:Change of the Orbital Ordering by Pressure
- Nonlinear Magnetic-Susceptibility of Two-Diemensional Magnets (C_nH_NH_3)_2CuCl_4 with n=1,2 and 3
- Reexamination of Magnetic Susceptibility and Spin Flop at Low-Field in KCuF_3
- Preparation and Magnetic Susceptibility of Cs_2CuF_4 and Rb_2CuF_4
- Efficient Blind MAI Suppression in DS/CDMA Systems by Embedded Constraint Parallel Projection Techniques(Digital Signal Processing, Papers Selected from the 19th Symposium on Signal Processing)
- Counting the Numbers of DOAs in Specified Ranges by Algebraic Phase Unwrapping Algorithms
- Formation of One-Dimensional Antiferromagnet in KCuF_3 with the Perovskite Structure
- Muon Spin Relaxation Study of Crossover from S=1/2 to S=1 of Composite Haldane Spin Chains in IPA-CuCl_3(Condensed matter: electronic structure and electrical, magnetic, and optical properties)
- A Note on Robust Adaptive Volterra Filtering Based on Parallel Subgradient Projection Techniques(Digital Signal Processing)
- New Superlattice Crystal Structure in KCuF_3 Revealed by X-Ray Diffraction Experiments
- Temperature and Frequency Dependence of the ESR Line Width in Two-Dimensional Ferromagnet K_2CuF_4
- A-4-10 Fast Stereo Echo Canceler by Adaptive Projected Subgradient Method with Multiple a priori Information
- Multi-Domain Adaptive Learning Based on Feasibility Splitting and Adaptive Projected Subgradient Method
- Formation of Alternating Antiferromagnetic Heisenberg Chains in (CH_3)_2CHNH_3CuBr_3
- Dimensional reduction techniques for adaptive generalized minor subspace estimation (通信方式)
- Dimensional reduction techniques for adaptive generalized minor subspace estimation (回路とシステム)
- Dimensional reduction techniques for adaptive generalized minor subspace estimation (信号処理)
- Two improvements of the fast iterative shrinkage/thresholding algorithm : Overrelaxation and weak convergence
- Mn-Implanted ZnS Thin Film Electroluminescent Cell
- Generalizing the multiple measurement setting from sparse vector recovery to low-rank matrix recovery (無線通信システム)
- Two improvements of the fast iterative shrinkage/thresholding algorithm: overrelaxation and weak convergence (通信方式)
- Magnetization of Two-Dimensional Ferromagnet K_2CuF_4
- NMR Study of Ferromagnetic K_2CuF_4
- Dynamic Effect of Low-Symmetric Distribution of Spins on EPR Lines in One-Dimensional Magnetic Systems. I. Theory and Experiment in TMMC
- Effect of Low-Symmetric Spin Distribution on EPR Lines in One-Dimensional Magnetic Systems
- Two improvements of the fast iterative shrinkage/thresholding algorithm: overrelaxation and weak convergence (回路とシステム)
- Generalizing the multiple measurement setting from sparse vector recovery to low-rank matrix recovery (信号処理)
- Observation of the Hyperfine Interaction in K_2CuF_4
- Multi-Sublattice Magnetic Structure of KCuF_3 Caused by the Antisymmetric Exchange Interaction : Antiferromagnetic Resonance Measurements
- Ferromagnetism in K_2CuF_4
- Convex Feasibility Problem with Prioritized Hard Constraints : Double Layered Projected Gradient Method(Numerical Analysis and Optimization)
- Shift of Half-Field EPR Lines in Low-Dimensional Magnetic Systems
- The Appearance of an Interdependent Effect of the Antisymmetric Exchange Interaction and Inequivalent g-Tensors in the Antiferromagnetic Resonance Field of KCuF_3
- Dynamic Effect of Low-Symmetric Distribution of Spins on EPR Lines in One-Dimensional Magnetic Systems. II. Enhancement by Cu Doping in TMMC
- A Visually Better Recovered Image Selection for Imaging Inverse Problems
- A Convergence Analysis of G-NPAST Algorithm for Extracting the First Principal Generalized Eigenvector
- Adaptive Generalized Eigen-pair Extraction by A Combination of Normalization and Quasi-Newton Algorithm
- Adaptive Generalized Eigen-pair Extraction by A Combination of Normalization and Quasi-Newton Algorithm
- Three variants of alternating direction method of multipliers without certain inner iterations and their application to image super-resolution via sparse representation
- Three variants of alternating direction method of multipliers without certain inner iterations and their application to image super-resolution via sparse representation
- Three variants of alternating direction method of multipliers without certain inner iterations and their application to image super-resolution via sparse representation
- Mixed Trigonometric Interpolation Techniques for Fast and Stable Algebraic Phase Unwrapping
- Adaptive proximal forward-backward splitting applied to Huber loss function for sparse system identification under impulsive noise
- A Visually Better Recovered Image Selection for Imaging Inverse Problems
- A Convergence Analysis of G-NPAST Algorithm for Extracting the First Principal Generalized Eigenvector
- Adaptive proximal forward-backward splitting applied to Huber loss function for sparse system identification under impulsive noise
- A Visually Better Recovered Image Selection for Imaging Inverse Problems
- A Convergence Analysis of G-NPAST Algorithm for Extracting the First Principal Generalized Eigenvector
- Adaptive proximal forward-backward splitting applied to Huber loss function for sparse system identification under impulsive noise
- Adaptive Generalized Eigen-pair Extraction by A Combination of Normalization and Quasi-Newton Algorithm
- Optical Birefringence Measurements in Ferromagnetic and Antiferromagnetic Alternating Heisenberg Chains (CH3)2CHNH3CuCl3
- Phonon-Raman Scattering in the Two-Dimensional Diluted System K_2Cu_xZn_F_4
- Tracking Analysis of Adaptive Filters with Data Normalization and Error Nonlinearities
- A Robust Algebraic Phase Unwrapping Based on Spline Approximation (システム数理と応用)
- A Robust Algebraic Phase Unwrapping Based on Spline Approximation (回路とシステム)
- Addendum to ESR Study of the Triangular Lattice Antiferromagnets with $S{=}1/2$: NaTiO2 and LiNiO2
- Detecting Reflection from Single Image Based on Nuclear Norm Minimization
- Sparsity-Aware Feed-Forward Active Noise Control with the Adaptive Douglas-Rachford Splitting
- Sparsity-Aware Feed-Forward Active Noise Control with the Adaptive Douglas-Rachford Splitting
- Detecting Reflection from Single Image Based on Nuclear Norm Minimization
- Detecting Reflection from Single Image Based on Nuclear Norm Minimization
- Sparsity-Aware Feed-Forward Active Noise Control with the Adaptive Douglas-Rachford Splitting
- Blockwise Low-Rank Prior for Cartoon-Texture Image Decomposition
- Blockwise Low-Rank Prior for Cartoon-Texture Image Decomposition
- Overlapped Local Color Nuclear Norm And Its Constrained Use
- Adaptive proximal forward-backward splitting applied to Huber loss function for sparse system identification under impulsive noise
- Adaptive proximal forward-backward splitting applied to Huber loss function for sparse system identification under impulsive noise
- Three variants of alternating direction method of multipliers without certain inner iterations and their application to image super-resolution via sparse representation