TY - GEN
T1 - Tarsier
T2 - 25th International Conference on Pattern Recognition
AU - Roziere, Baptiste
AU - Rakotonirina, Nathanaël Carraz
AU - Hosu, Vlad
AU - Rasoanaivo, Andry
AU - Lin, Hanhe
AU - Couprie, Camille
AU - Teytaud, Olivier
N1 - Funding Information:
Vlad Hosu and Hanhe Lin from the University of Konstanz were funded by the Deutsche Forschungsgemeinschaft (DFG), Project-ID 251654672, TRR 161 (Project A05).
Publisher Copyright:
© 2020 IEEE
PY - 2020
Y1 - 2020
N2 - 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.
AB - 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.
UR - http://www.scopus.com/inward/record.url?scp=85106727461&partnerID=8YFLogxK
U2 - 10.1109/ICPR48806.2021.9413318
DO - 10.1109/ICPR48806.2021.9413318
M3 - Conference contribution
AN - SCOPUS:85106727461
SN - 978-1-7281-8809-6
T3 - Proceedings - International Conference on Pattern Recognition
SP - 7028
EP - 7035
BT - 2020 25th International Conference on Pattern Recognition (ICPR)
PB - IEEE
Y2 - 10 January 2021 through 15 January 2021
ER -