GAN with Pixel and Perceptual Regularizations for Photo-Realistic Joint Deblurring and Super-Resolution

Research output: Chapters, Conference Papers, Creative and Literary Works (RGC: 12, 32, 41, 45)32_Refereed conference paper (with host publication)peer-review

2 Scopus Citations
View graph of relations

Author(s)

  • Yong Li
  • Zhenguo Yang
  • Yong Wang
  • Qing Li
  • Wenyin Liu
  • Ying Wang

Related Research Unit(s)

Detail(s)

Original languageEnglish
Title of host publicationAdvances in Computer Graphics
Subtitle of host publicationProceedings
EditorsMarina Gavrilova, Jian Chang, Nadia Magnenat Thalmann, Eckhard Hitzer, Hiroshi Ishikawa
PublisherSpringer Nature Switzerland AG
Pages395-401
ISBN (Electronic)9783030225148
ISBN (Print)9783030225131
Publication statusPublished - Jun 2019

Publication series

NameLecture Notes in Computer Science
Volume11542 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Title36th Computer Graphics International Conference (CGI 2019)
LocationUniversity of Calgary
PlaceCanada
CityCalgary
Period17 - 20 June 2019

Abstract

In this paper, we propose a Generative Adversarial Network with Pixel and Perceptual regularizations, denoted as P2GAN, to restore single motion blurry and low-resolution images jointly into clear and high-resolution images. It is an end-to-end neural network consisting of deblurring module and super-resolution module, which repairs degraded pixels in the motion-blur images firstly, and then outputs the deblurred images and deblurred features for further reconstruction. More specifically, the proposed P2GAN integrates pixel-wise loss in pixel-level, contextual loss and adversarial loss in perceptual level simultaneously, in order to guide on deblurring and super-resolution reconstruction of the raw images that are blurry and in low-resolution, which help obtaining realistic images. Extensive experiments conducted on a real-world dataset manifest the effectiveness of the proposed approaches, outperforming the state-of-the-art models.

Research Area(s)

  • Contextual loss, GANs, Image deblurring, Pixel loss, Super-resolution

Citation Format(s)

GAN with Pixel and Perceptual Regularizations for Photo-Realistic Joint Deblurring and Super-Resolution. / Li, Yong; Yang, Zhenguo; Mao, Xudong et al.
Advances in Computer Graphics: Proceedings. ed. / Marina Gavrilova; Jian Chang; Nadia Magnenat Thalmann; Eckhard Hitzer; Hiroshi Ishikawa. Springer Nature Switzerland AG, 2019. p. 395-401 (Lecture Notes in Computer Science; Vol. 11542 LNCS).

Research output: Chapters, Conference Papers, Creative and Literary Works (RGC: 12, 32, 41, 45)32_Refereed conference paper (with host publication)peer-review