English

Structure-Preserving Super Resolution with Gradient Guidance

Image and Video Processing 2020-03-31 v1 Computer Vision and Pattern Recognition

Abstract

Structures matter in single image super resolution (SISR). Recent studies benefiting from generative adversarial network (GAN) have promoted the development of SISR by recovering photo-realistic images. However, there are always undesired structural distortions in the recovered images. In this paper, we propose a structure-preserving super resolution method to alleviate the above issue while maintaining the merits of GAN-based methods to generate perceptual-pleasant details. Specifically, we exploit gradient maps of images to guide the recovery in two aspects. On the one hand, we restore high-resolution gradient maps by a gradient branch to provide additional structure priors for the SR process. On the other hand, we propose a gradient loss which imposes a second-order restriction on the super-resolved images. Along with the previous image-space loss functions, the gradient-space objectives help generative networks concentrate more on geometric structures. Moreover, our method is model-agnostic, which can be potentially used for off-the-shelf SR networks. Experimental results show that we achieve the best PI and LPIPS performance and meanwhile comparable PSNR and SSIM compared with state-of-the-art perceptual-driven SR methods. Visual results demonstrate our superiority in restoring structures while generating natural SR images.

Keywords

Cite

@article{arxiv.2003.13081,
  title  = {Structure-Preserving Super Resolution with Gradient Guidance},
  author = {Cheng Ma and Yongming Rao and Yean Cheng and Ce Chen and Jiwen Lu and Jie Zhou},
  journal= {arXiv preprint arXiv:2003.13081},
  year   = {2020}
}

Comments

Accepted to CVPR 2020

R2 v1 2026-06-23T14:30:59.273Z