Valkonen, Tuomo 
ORCID: 0000-0001-6683-3572 and Pock, Thomas
(2017)
Acceleration of the PDHGM on Partially Strongly Convex Functions.
JOURNAL OF MATHEMATICAL IMAGING AND VISION, 59 (3).
pp. 394-414.
![[img]](https://livrepository.liverpool.ac.uk/style/images/fileicons/application_atom+xml.png) |
Atom XML (admin)
2017-05-12T11:29:22Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-12T12:25:48Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-12T13:25:01Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-12T14:25:04Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-12T15:25:14Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-12T16:25:40Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-12T17:25:05Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-12T18:25:55Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-12T19:25:18Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-12T20:25:38Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-12T21:26:00Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-12T22:26:48Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-12T23:32:21Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-13T01:27:13Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-13T02:26:40Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-13T03:26:47Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-13T04:49:23Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-13T05:26:42Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-13T06:25:52Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-13T07:26:10Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-13T08:26:35Z.atom - Unspecified Download (0B) |
|
Atom XML (admin)
2017-05-13T09:25:47Z.atom - Unspecified Download (0B) |
Abstract
We propose several variants of the primal-dual method due to Chambolle and Pock. Without requiring full strong convexity of the objective functions, our methods are accelerated on subspaces with strong convexity. This yields mixed rates, O(1/N2) with respect to initialisation and <i>O</i>(1 / <i>N</i>) with respect to the dual sequence, and the residual part of the primal sequence. We demonstrate the efficacy of the proposed methods on image processing problems lacking strong convexity, such as total generalised variation denoising and total variation deblurring.
| Item Type: | Article |
|---|---|
| Uncontrolled Keywords: | Primal-dual, Accelerated, Subspace, Total generalised variation |
| Depositing User: | Symplectic Admin |
| Date Deposited: | 25 Apr 2017 06:37 |
| Last Modified: | 02 Apr 2024 09:25 |
| DOI: | 10.1007/s10851-016-0692-2 |
| Open Access URL: | http://link.springer.com/article/10.1007/s10851-01... |
| Related URLs: | |
| URI: | https://livrepository.liverpool.ac.uk/id/eprint/3007115 |
Dimensions
Dimensions
