Total Deep Variation for Linear Inverse Problems

MaRDI QIDQ6332892arXivFDO

Authors Erich Kobler, Alexander Effland, Karl Kunisch, Thomas Pock

Publication date 14 January 2020

Published in arXiv

Has companion code repository https://github.com/VLOGroup/tdv

Mathematics Subject Classification ID

Control problems involving ordinary differential equations (34H05) Optimality conditions for problems involving ordinary differential equations (49K15) Numerical methods for initial value problems involving ordinary differential equations (65L05) Machine vision and scene understanding (68T45)

arXiv classification

cs.CVcs.LGmath.OC

Abstract: Diverse inverse problems in imaging can be cast as variational problems composed of a task-specific data fidelity term and a regularization term. In this paper, we propose a novel learnable general-purpose regularizer exploiting recent architectural design patterns from deep learning. We cast the learning problem as a discrete sampled optimal control problem, for which we derive the adjoint state equations and an optimality condition. By exploiting the variational structure of our approach, we perform a sensitivity analysis with respect to the learned parameters obtained from different training datasets. Moreover, we carry out a nonlinear eigenfunction analysis, which reveals interesting properties of the learned regularizer. We show state-of-the-art performance for classical image restoration and medical image reconstruction problems.

Cited in

(1)

• Total Generalized Variation for Inverse Problems Involving Piecewise Linear Finite Elements

This page was built for publication: Total Deep Variation for Linear Inverse Problems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6332892)