Toward Real-World Single Image Super-Resolution: A New Benchmark and A New Model

被引:385
作者
Cai, Jianrui [1 ]
Zeng, Hui [1 ]
Yong, Hongwei [1 ,3 ]
Cao, Zisheng [2 ]
Zhang, Lei [1 ,3 ]
机构
[1] Hong Kong Polytech Univ, Hong Kong, Peoples R China
[2] DJI Co Ltd, Shenzhen, Guangdong, Peoples R China
[3] Alibaba Grp, DAMO Acad, Hangzhou, Zhejiang, Peoples R China
来源
2019 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2019) | 2019年
关键词
D O I
10.1109/ICCV.2019.00318
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Most of the existing learning-based single image super-resolution (SISR) methods are trained and evaluated on simulated datasets, where the low-resolution (LR) images are generated by applying a simple and uniform degradation (i.e., bicubic downsampling) to their high-resolution (HR) counterparts. However, the degradations in real-world LR images are far more complicated. As a consequence, the SISR models trained on simulated data become less effective when applied to practical scenarios. In this paper, we build a real-world super-resolution (RealSR) dataset where paired LR-HR images on the same scene are captured by adjusting the focal length of a digital camera. An image registration algorithm is developed to progressively align the image pairs at different resolutions. Considering that the degradation kernels are naturally non-uniform in our dataset, we present a Laplacian pyramid based kernel prediction network (LP-KPN), which efficiently learns per-pixel kernels to recover the HR image. Our extensive experiments demonstrate that SISR models trained on our RealSR dataset deliver better visual quality with sharper edges and finer textures on real-world scenes than those trained on simulated datasets. Though our RealSR dataset is built by using only two cameras (Canon 5D3 and Nikon D810), the trained model generalizes well to other camera devices such as Sony a7II and mobile phones.
引用
收藏
页码:3086 / 3095
页数:10
相关论文
共 63 条
[1]  
Agustsson Eirikur, 2017, CVPRW
[2]  
[Anonymous], 2014, ACM MM
[3]  
[Anonymous], 2018, CVPR
[4]  
[Anonymous], 2016, ECCV
[5]  
[Anonymous], 2012, LOW COMPLEXITY SINGL
[6]  
[Anonymous], 2004, CVPR
[7]  
[Anonymous], 2017, ICCV
[8]  
[Anonymous], 2013, ICCV
[9]  
[Anonymous], 2015, Nature, DOI [10.1038/nature14539, DOI 10.1038/NATURE14539]
[10]  
[Anonymous], 2008, ICASSP