Tarsier: evolving noise injection in super-resolution GANs

Baptiste Roziere, Nathanaël Carraz Rakotonirina, Vlad Hosu, Andry Rasoanaivo, Hanhe Lin, Camille Couprie, Olivier Teytaud

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

Super-resolution aims at increasing the resolution and level of detail within an image. The current state of the art in general single-image super-resolution is held by NESRGAN+, which injects a Gaussian noise after each residual layer at training time. In this paper, we harness evolutionary methods to improve NESRGAN+ by optimizing the noise injection at inference time. More precisely, we use Diagonal CMA to optimize the injected noise according to a novel criterion combining quality assessment and realism. Our results are validated by the PIRM perceptual score and a human study. Our method outperforms NESRGAN+ on several standard super-resolution datasets. More generally, our approach can be used to optimize any method based on noise injection.

Original languageEnglish
Title of host publication2020 25th International Conference on Pattern Recognition (ICPR)
PublisherIEEE
Pages7028-7035
Number of pages8
ISBN (Electronic)978-1-7281-8808-9
ISBN (Print)978-1-7281-8809-6
DOIs
Publication statusPublished - 2020
Event25th International Conference on Pattern Recognition - Virtual, Milan, Italy
Duration: 10 Jan 202115 Jan 2021

Publication series

NameProceedings - International Conference on Pattern Recognition
PublisherIEEE
ISSN (Print)1051-4651

Conference

Conference25th International Conference on Pattern Recognition
Abbreviated titleICPR 2020
Country/TerritoryItaly
CityVirtual, Milan
Period10/01/2115/01/21

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition

Fingerprint

Dive into the research topics of 'Tarsier: evolving noise injection in super-resolution GANs'. Together they form a unique fingerprint.

Cite this