author：
Lugmayr, Andreas
(Lugmayr, Andreas.)
[1]
|
Danelljan, Martin
(Danelljan, Martin.)
[2]
|
Timofte, Radu
(Timofte, Radu.)
[3]
|
Ahn, Namhyuk
(Ahn, Namhyuk.)
[4]
|
Bai, Dongwoon
(Bai, Dongwoon.)
[5]
|
Cai, Jie
(Cai, Jie.)
[6]
|
Cao, Yun
(Cao, Yun.)
[7]
|
Chen, Junyang
(Chen, Junyang.)
[8]
|
Cheng, Kaihua
(Cheng, Kaihua.)
[9]
|
Chun, Seyoung
(Chun, Seyoung.)
[10]
|
Deng, Wei
(Deng, Wei.)
[11]
|
El-Khamy, Mostafa
(El-Khamy, Mostafa.)
[12]
|
Ho, Chiu Man
(Ho, Chiu Man.)
[13]
|
Ji, Xiaozhong
(Ji, Xiaozhong.)
[14]
|
Kheradmand, Amin
(Kheradmand, Amin.)
[15]
|
Kim, Gwantae
(Kim, Gwantae.)
[16]
|
Ko, Hanseok
(Ko, Hanseok.)
[17]
|
Lee, Kanghyu
(Lee, Kanghyu.)
[18]
|
Lee, Jungwon
(Lee, Jungwon.)
[19]
|
Li, Hao
(Li, Hao.)
[20]
|
Liu, Ziluan
(Liu, Ziluan.)
[21]
|
Liu, Zhi-Song
(Liu, Zhi-Song.)
[22]
|
Liu, Shuai
(Liu, Shuai.)
[23]
|
Lu, Yunhua
(Lu, Yunhua.)
[24]
|
Meng, Zibo
(Meng, Zibo.)
[25]
|
Michelini, Pablo Navarrete
(Michelini, Pablo Navarrete.)
[26]
|
Micheloni, Christian
(Micheloni, Christian.)
[27]
|
Prajapati, Kalpesh
(Prajapati, Kalpesh.)
[28]
|
Ren, Haoyu
(Ren, Haoyu.)
[29]
|
Seo, Yong Hyeok
(Seo, Yong Hyeok.)
[30]
|
Siu, Wan-Chi
(Siu, Wan-Chi.)
[31]
|
Sohn, Kyung-Ah
(Sohn, Kyung-Ah.)
[32]
|
Tai, Ying
(Tai, Ying.)
[33]
|
Umer, Rao Muhammad
(Umer, Rao Muhammad.)
[34]
|
Wang, Shuangquan
(Wang, Shuangquan.)
[35]
|
Wang, Huibing
(Wang, Huibing.)
[36]
|
Wu, Timothy Haoning
(Wu, Timothy Haoning.)
[37]
|
Wu, Haoning
(Wu, Haoning.)
[38]
|
Yang, Biao
(Yang, Biao.)
[39]
|
Yang, Fuzhi
(Yang, Fuzhi.)
[40]
|
Yoo, Jaejun
(Yoo, Jaejun.)
[41]
|
Zhao, Tongtong
(Zhao, Tongtong.)
[42]
|
Zhou, Yuanbo
(Zhou, Yuanbo.)
[43]
|
Zhuo, Haijie
(Zhuo, Haijie.)
[44]
|
Zong, Ziyao
(Zong, Ziyao.)
[45]
|
Zou, X.
(Zou, X..)
[46]
Abstract：
This paper reviews the NTIRE 2020 challenge on real world super-resolution. It focuses on the participating methods and final results. The challenge addresses the real world setting, where paired true high and low-resolution images are unavailable. For training, only one set of source input images is therefore provided along with a set of unpaired high-quality target images. In Track 1: Image Processing artifacts, the aim is to super-resolve images with synthetically generated image processing artifacts. This allows for quantitative benchmarking of the approaches w.r.t. a ground-truth image. In Track 2: Smartphone Images, real low-quality smart phone images have to be super-resolved. In both tracks, the ultimate goal is to achieve the best perceptual quality, evaluated using a human study. This is the second challenge on the subject, following AIM 2019, targeting to advance the state-of-the-art in super-resolution. To measure the performance we use the benchmark protocol from AIM 2019. In total 22 teams competed in the final testing phase, demonstrating new and innovative solutions to the problem. © 2020 IEEE.
Keyword：
Benchmarking
Computer vision
Optical resolving power
Smartphones
