Regularization by Architecture: A Deep Prior Approach for Inverse Problems

被引：61

作者：

Dittmer, Soren ^{[1
]}

Kluth, Tobias ^{[1
]}

Maass, Peter ^{[1
]}

Baguer, Daniel Otero ^{[1
]}

机构：

[1] Univ Bremen, Ctr Ind Math ZeTeM, Bremen, Germany

来源：

JOURNAL OF MATHEMATICAL IMAGING AND VISION | 2020年 / 62卷 / 03期

基金：

欧盟地平线“2020”;

关键词：

Inverse problems; Deep learning; Regularization by architecture; Deep inverse prior; Deep image prior; THRESHOLDING ALGORITHM; SPARSE;

D O I：

10.1007/s10851-019-00923-x

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

The present paper studies so-called deep image prior (DIP) techniques in the context of ill-posed inverse problems. DIP networks have been recently introduced for applications in image processing; also first experimental results for applying DIP to inverse problems have been reported. This paper aims at discussing different interpretations of DIP and to obtain analytic results for specific network designs and linear operators. The main contribution is to introduce the idea of viewing these approaches as the optimization of Tikhonov functionals rather than optimizing networks. Besides theoretical results, we present numerical verifications.

引用

页码：456 / 470

页数：15

共 50 条

[1] Regularization by Architecture: A Deep Prior Approach for Inverse Problems
Sören Dittmer
Tobias Kluth
Peter Maass
Daniel Otero Baguer
[J]. Journal of Mathematical Imaging and Vision, 2020, 62 : 456 - 470
[2] Regularization theory of the analytic deep prior approach
Arndt, Clemens
[J]. INVERSE PROBLEMS, 2022, 38 (11)
[3] Incremental projection approach of regularization for inverse problems
Souopgui, Innocent
Ngodock, Hans E.
Vidard, Arthur
Le Dimet, Francois-Xavier
[J]. APPLIED MATHEMATICS AND OPTIMIZATION, 2016, 74 (02): : 303 - 324
[4] Incremental projection approach of regularization for inverse problems
Innocent Souopgui
Hans E. Ngodock
Arthur Vidard
François-Xavier Le Dimet
[J]. Applied Mathematics & Optimization, 2016, 74 : 303 - 324
[5] Deep unfolding as iterative regularization for imaging inverse problems
Cui, Zhuo-Xu
Zhu, Qingyong
Cheng, Jing
Zhang, Bo
Liang, Dong
[J]. INVERSE PROBLEMS, 2024, 40 (02)
[6] Deep image prior and weighted anisotropic-isotropic total variation regularization for solving linear inverse problems
Xie, Yujia
Chen, Wengu
Ge, Huanmin
Ng, Michael K.
[J]. APPLIED MATHEMATICS AND COMPUTATION, 2024, 482
[7] REGULARIZATION AND INVERSE PROBLEMS
SIBISI, S
[J]. MAXIMUM ENTROPY AND BAYESIAN METHODS /, 1989, 36 : 389 - 396
[8] Regularization and inverse problems
Lasenby, A
Barreiro, B
Hobson, M
[J]. MINING THE SKY, 2001, : 15 - 32
[9] Total Deep Variation: A Stable Regularization Method for Inverse Problems
Kobler, Erich
Effland, Alexander
Kunisch, Karl
Pock, Thomas
[J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2022, 44 (12) : 9163 - 9180
[10] REGULARIZATION OF INVERSE PROBLEMS
ROMANOVSKII, MR
[J]. HIGH TEMPERATURE, 1980, 18 (01) : 135 - 140

← 1 2 3 4 5 →