Paper Info
Title
GAN with Pixel and Perceptual Regularizations for Photo-Realistic Joint Deblurring and Super-Resolution.
Abstract
In this paper, we propose a Generative Adversarial Network with Pixel and Perceptual regularizations, denoted as P(2)GAN, 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 P(2)GAN 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.
Year
DOI
Venue
2019
10.1007/978-3-030-22514-8_36
ADVANCES IN COMPUTER GRAPHICS, CGI 2019
Keywords
Field
DocType
Image deblurring,Super-resolution,GANs,Pixel loss,Contextual loss
Computer vision,Generative adversarial network,Deblurring,Computer science,Artificial intelligence,Pixel,Artificial neural network,Superresolution,Perception
Conference
Volume
ISSN
Citations 
11542
0302-9743
0
PageRank 
References 
Authors
0.34
0
7
Name
Order
Citations
PageRank
Yong Li133.66
Zhenguo Yang27117.57
Xudong Mao310510.64
Yong Wang4425.11
Qing Li53222433.87
Liu Wenyin62531215.13
Ying Wang700.34