Scale-recurrent Network for Deep Image Deblurring

In single image deblurring, the "coarse-to-fine" scheme, i.e. gradually restoring the sharp image on different resolutions in a pyramid, is very successful in both traditional optimization-based methods and recent neural-network-based approaches. In this paper, we investigate this strategy and propose a Scale-recurrent Network (SRN-DeblurNet) for this deblurring task. Compared with the many recent learning-based approaches in [25], it has a simpler network structure, a smaller number of parameters and is easier to train. We evaluate our method on large-scale deblurring datasets with complex motion. Results show that our method can produce better quality results than state-of-the-arts, both quantitatively and qualitatively.

PDF Abstract CVPR 2018 PDF CVPR 2018 Abstract

Results from the Paper

Task Dataset Model Metric Name Metric Value Global Rank Uses Extra
Training Data
Result Benchmark
Deblurring GoPro SRN SSIM 0.9342 # 22
Image Deblurring GoPro SRN SSIM 0.9342 # 20
Deblurring HIDE (trained on GOPRO) SRN PSNR (sRGB) 28.36 # 12
SSIM (sRGB) 0.915 # 10
Deblurring RealBlur-J SRN SSIM (sRGB) 0.909 # 5
PSNR (sRGB) 31.38 # 5
Deblurring RealBlur-J (trained on GoPro) SRN PSNR (sRGB) 28.56 # 7
Deblurring RealBlur-R SRN PSNR (sRGB) 38.65 # 4
SSIM (sRGB) 0.965 # 3
Deblurring RealBlur-R (trained on GoPro) SRN SSIM (sRGB) 0.947 # 6
Image Relighting VIDIT’20 validation set SRN PSNR 16.94 # 5
SSIM 0.5660 # 5
LPIPS 0.4319 # 5
MPS 0.5670 # 6
Runtime(s) 0.87 # 5


No methods listed for this paper. Add relevant methods here