Deep Back-Projection Networks for Super-Resolution

The feed-forward architectures of recently proposed deep super-resolution networks learn representations of low-resolution inputs, and the non-linear mapping from those to high-resolution output. However, this approach does not fully address the mutual dependencies of low- and high-resolution images...

Full description

Saved in:
Bibliographic Details
Published in2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition pp. 1664 - 1673
Main Authors Haris, Muhammad, Shakhnarovich, Greg, Ukita, Norimichi
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.06.2018
Subjects
Computer architecture
Feature extraction
Image reconstruction
Spatial resolution
Task analysis
Training
Online AccessGet full text

Cover

More Information
Summary:The feed-forward architectures of recently proposed deep super-resolution networks learn representations of low-resolution inputs, and the non-linear mapping from those to high-resolution output. However, this approach does not fully address the mutual dependencies of low- and high-resolution images. We propose Deep Back-Projection Networks (DBPN), that exploit iterative up- and downsampling layers, providing an error feedback mechanism for projection errors at each stage. We construct mutually-connected up- and down-sampling stages each of which represents different types of image degradation and high-resolution components. We show that extending this idea to allow concatenation of features across up- and downsampling stages (Dense DBPN) allows us to reconstruct further improve super-resolution, yielding superior results and in particular establishing new state of the art results for large scaling factors such as 8Ã- across multiple data sets.
ISSN:1063-6919
DOI:10.1109/CVPR.2018.00179