A First-Order Primal-Dual Algorithm for Convex Problems with Applications to Imaging

被引:2976
作者
Chambolle, Antonin [1 ]
Pock, Thomas [2 ]
机构
[1] Ecole Polytech, CNRS, CMAP, F-91128 Palaiseau, France
[2] Graz Univ Technol, Inst Comp Graph & Vis, A-8010 Graz, Austria
基金
奥地利科学基金会;
关键词
Convex optimization; Dual approaches; Total variation; Inverse problems; Image; Reconstruction; MONOTONE-OPERATORS; SOAP FILMS; POINT; MINIMIZATION;
D O I
10.1007/s10851-010-0251-1
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In this paper we study a first-order primal-dual algorithm for non-smooth convex optimization problems with known saddle-point structure. We prove convergence to a saddle-point with rate O(1/N) in finite dimensions for the complete class of problems. We further show accelerations of the proposed algorithm to yield improved rates on problems with some degree of smoothness. In particular we show that we can achieve O(1/N (2)) convergence on problems, where the primal or the dual objective is uniformly convex, and we can show linear convergence, i.e. O(omega (N) ) for some omega a(0,1), on smooth problems. The wide applicability of the proposed algorithm is demonstrated on several imaging problems such as image denoising, image deconvolution, image inpainting, motion estimation and multi-label image segmentation.
引用
收藏
页码:120 / 145
页数:26
相关论文
共 38 条
[31]  
Pock T., 2009, LNCS
[32]  
Popov L.D., 1980, MAT ZAMETKI, V28, P803
[33]  
Rockafellar R. T., 1970, Convex analysis
[34]   MONOTONE OPERATORS AND PROXIMAL POINT ALGORITHM [J].
ROCKAFELLAR, RT .
SIAM JOURNAL ON CONTROL, 1976, 14 (05) :877-898
[35]   NONLINEAR TOTAL VARIATION BASED NOISE REMOVAL ALGORITHMS [J].
RUDIN, LI ;
OSHER, S ;
FATEMI, E .
PHYSICA D, 1992, 60 (1-4) :259-268
[36]  
Shulman D., 1989, Proceedings. Workshop on Visual Motion (IEEE Cat. No.89CH2716-9), P81, DOI 10.1109/WVM.1989.47097
[37]   A duality based approach for realtime TV-L1 optical flow [J].
Zach, C. ;
Pock, T. ;
Bischof, H. .
PATTERN RECOGNITION, PROCEEDINGS, 2007, 4713 :214-+
[38]  
ZHU M, 2008, CAM REPORTS, V834