Document Type

Journal Article

Department/Unit

Department of Mathematics

Title

On the convergence of primal-dual hybrid gradient algorithm

Language

English

Abstract

© 2014 Society for Industrial and Applied Mathematics. The primal-dual hybrid gradient algorithm (PDHG) has been widely used, especially for some basic image processing models. In the literature, PDHG’s convergence was established only under some restrictive conditions on its step sizes. In this paper, we revisit PDHG’s convergence in the context of a saddle-point problem and try to better understand how to choose its step sizes. More specifically, we show by an extremely simple example that PDHG is not necessarily convergent even when the step sizes are fixed as tiny constants. We then show that PDHG with constant step sizes is indeed convergent if one of the functions of the saddle-point problem is strongly convex, a condition that does hold for some variational models in imaging. With this additional condition, we also establish a worst-case convergence rate measured by the iteration complexity for PDHG with constant step sizes.

Keywords

Convergence rate, Convex optimization, Image restoration, Primal-dual hybrid gradient algorithm, Saddle-point problem, Total variation

Publication Date

2014

Source Publication Title

SIAM Journal on Imaging Sciences

Volume

7

Issue

4

Start Page

2526

End Page

2537

Publisher

Society for Industrial and Applied Mathematics

DOI

10.1137/140963467

Link to Publisher's Edition

http://dx.doi.org/10.1137/140963467

ISSN (electronic)

19364954

This document is currently not available here.

Link to Publisher's Version

Share

COinS