Corpus ID: 231839458

Real-World Super-Resolution of Face-Images from Surveillance Cameras

@article{Aakerberg2021RealWorldSO,
  title={Real-World Super-Resolution of Face-Images from Surveillance Cameras},
  author={Andreas Aakerberg and Kamal Nasrollahi and T. Moeslund},
  journal={ArXiv},
  year={2021},
  volume={abs/2102.03113}
}
Most existing face image Super-Resolution (SR) methods assume that the Low-Resolution (LR) images were artificially downsampled from High-Resolution (HR) images with bicubic interpolation. This operation changes the natural image characteristics and reduces noise. Hence, SR methods trained on such data most often fail to produce good results when applied to real LR images. To solve this problem, we propose a novel framework for generation of realistic LR/HR training pairs. Our framework… Expand

Figures and Tables from this paper

References

SHOWING 1-10 OF 60 REFERENCES
Real-World Super-Resolution via Kernel Estimation and Noise Injection
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network
  • C. Ledig, Lucas Theis, +6 authors W. Shi
  • Computer Science, Mathematics
  • 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2017
Toward Real-World Single Image Super-Resolution: A New Benchmark and a New Model
Frequency Separation for Real-World Super-Resolution
Blind Super-Resolution Kernel Estimation using an Internal-GAN
Characteristic Regularisation for Super-Resolving Face Images
ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks
Unsupervised Learning for Real-World Super-Resolution
ImagePairs: Realistic Super Resolution Dataset via Beam Splitter Camera Rig
...
1
2
3
4
5
...