The Learned Inexact Project Gradient Descent Algorithm

Conference Paper

Accelerating iterative algorithms for solving inverse problems using neural networks have become a very popular strategy in the recent years. In this work, we propose a theoretical analysis that may provide an explanation for its success. Our theory relies on the usage of inexact projections with the projected gradient descent (PGD) method. It is demonstrated in various problems including image super-resolution.

Full Text

Duke Authors

Cited Authors

  • Giryes, R; Eldar, YC; Bronstein, AM; Sapiro, G

Published Date

  • September 10, 2018

Published In

Volume / Issue

  • 2018-April /

Start / End Page

  • 6767 - 6771

International Standard Serial Number (ISSN)

  • 1520-6149

International Standard Book Number 13 (ISBN-13)

  • 9781538646588

Digital Object Identifier (DOI)

  • 10.1109/ICASSP.2018.8462136

Citation Source

  • Scopus