Image super-resolution based on conditional generative adversarial network

Generative adversarial network (GAN) is one of the most prevalent generative models that can synthesise realistic high-frequency details. However, a mismatch between the input and the output may arise when GAN is directly applied to image super-resolution. To alleviate this issue, the authors adopte...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IET image processing Jg. 14; H. 13; S. 3006 - 3013
Hauptverfasser: Gao, Hongxia, Chen, Zhanhong, Huang, Binyang, Chen, Jiahe, Li, Zhifu
Format: Journal Article
Sprache:Englisch
Veröffentlicht: The Institution of Engineering and Technology 01.11.2020
Schlagworte:
adversarial loss term
cGAN discriminator
conditional GAN
conditional generative adversarial network
generator loss function
high‐frequency information
high‐resolution image
HR images
image resolution
image super‐resolution
L1 loss term
low‐frequency information
low‐resolution image
neural nets
Research Article
unsupervised learning
adversarial loss term
conditional GAN
high-frequency information
conditional generative adversarial network
high-resolution image
unsupervised learning
HR images
generator loss function
cGAN discriminator
L1 loss term
low-frequency information
image resolution
neural nets
image super-resolution
low-resolution image
ISSN:1751-9659, 1751-9667
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Generative adversarial network (GAN) is one of the most prevalent generative models that can synthesise realistic high-frequency details. However, a mismatch between the input and the output may arise when GAN is directly applied to image super-resolution. To alleviate this issue, the authors adopted a conditional GAN (cGAN) in this study. The cGAN discriminator attempted to guess whether the unknown high-resolution (HR) image was produced by the generator with the aid of the original low-resolution (LR) image. They propose a novel discriminator that only penalises at the scale of the patch and, thus, has relatively few parameters to train. The generator of cGAN is an encoder–decoder with skip connections to shuttle the shared low-level information directly across the network. To better maintain the low-frequency information and recover the high-frequency information, they designed a generator loss function combining adversarial loss term and L1 loss term. The former term is beneficial to the synthesis of fine-grained textures, while the latter is responsible for learning the overall structure of the LR input. The experiments revealed that the proposed method could generate HR images with richer details and less over-smoothness.
ISSN:1751-9659
1751-9667
DOI:10.1049/iet-ipr.2018.5767