Corpus ID: 202577523

Blind Super-Resolution Kernel Estimation using an Internal-GAN

@inproceedings{BellKligler2019BlindSK,
  title={Blind Super-Resolution Kernel Estimation using an Internal-GAN},
  author={Sefi Bell-Kligler and Assaf Shocher and Michal Irani},
  booktitle={NeurIPS},
  year={2019}
}
Super resolution (SR) methods typically assume that the low-resolution (LR) image was downscaled from the unknown high-resolution (HR) image by a fixed `ideal’ downscaling kernel (e.g. Bicubic downscaling). However, this is rarely the case in real LR images, in contrast to synthetically generated SR datasets. When the assumed downscaling kernel deviates from the true one, the performance of SR methods significantly deteriorates. This gave rise to Blind-SR - namely, SR when the downscaling… Expand

Paper Mentions

Blog Post
Dual Back-Projection-Based Internal Learning for Blind Super-Resolution
TLDR
This letter proposes a unified internal learning-based blind SR method that jointly trains two image-specific networks for 1) downscaling together with kernel estimation and 2) SR performance, and shows that the proposed method outperforms previous two-stage internallearning-based methods in terms of both SR performance and efficiency. Expand
KernelNet: A Blind Super-Resolution Kernel Estimation Network
TLDR
The proposed model outperforms the state-of-the-art SR kernel estimator, KernelGAN, by a significant margin in SR kernel reconstruction accuracy and is the first one that can estimate the SR kernel in real-time by performing O(1k) times faster than KernelGAN. Expand
Toward Real-World Super-Resolution via Adaptive Downsampling Models
  • Sanghyun Son, Jaeha Kim, Wei-Sheng Lai, Ming-Hsuan Yang, Kyoung Mu Lee
  • Engineering, Computer Science
  • IEEE transactions on pattern analysis and machine intelligence
  • 2021
TLDR
A novel method to simulate an unknown real downsampling process without imposing restrictive prior knowledge is proposed and an effective and generalizable low-frequency loss (LFL) in the adversarial training framework is designed to imitate the distribution of target LR images without using any paired examples. Expand
Frequency Consistent Adaptation for Real World Super Resolution
TLDR
A novel Frequency Consistent Adaptation (FCA) is designed that improves the performance of the SR model under real-world setting achieving state-of-the-art results with high fidelity and plausible perception, thus providing a novel effective framework for realworld SR application. Expand
Perceptual-based super-resolution reconstruction using image-specific degradation estimation
TLDR
Results show the effectiveness of the image-degradation estimation method in extracting inner-image statistics for a better super-resolution perceptual reconstruction. Expand
Learning cascaded convolutional networks for blind single image super-resolution
TLDR
This paper learns blind SR network from a realistic, parametric degradation model by considering blurring, noise, downsampling, and even JPEG compression, which can be jointly end-to-end learned from training data and benefit generalization ability. Expand
DynaVSR: Dynamic Adaptive Blind Video Super-Resolution
TLDR
DynaVSR is presented, a novel meta-learning-based framework for real-world video SR that enables efficient downscaling model estimation and adaptation to the current input and consistently improves the performance of the state-of-the-art video SR models by a large margin. Expand
Benefitting from Bicubically Down-Sampled Images for Learning Real-World Image Super-Resolution
TLDR
This work proposes to handle real-world SR by splitting this ill-posed problem into two comparatively more well-posed steps, and shows that the proposed approach outperforms the state-of-the-art works in terms of both qualitative and quantitative results. Expand
Temporal Kernel Consistency for Blind Video Super-Resolution
TLDR
The temporal properties of the kernels are investigated and its importance in the task of blind video super-resolution is highlighted, achieving the new state-of-the-art in blind video SR and underlining the potential of exploiting kernel temporal consistency. Expand
MRI super-resolution via realistic downsampling with adversarial learning
  • Bangyan Huang, Haonan Xiao, +8 authors Jing Cai
  • Physics, Medicine
  • Physics in medicine and biology
  • 2021
TLDR
This study proposed a DL-based SR framework with an emphasis on data construction to achieve better performance on real LR MR images and exhibited significantly better resolution enhancement and fewer artifacts compared with the other two networks. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 41 REFERENCES
Blind Super-Resolution With Iterative Kernel Correction
TLDR
An iterative correction scheme -- IKC that achieves better results than direct kernel estimation in blind SR problem and an effective SR network architecture using spatial feature transform (SFT) layers to handle multiple blur kernels, named SFTMD. Expand
"Zero-Shot" Super-Resolution Using Deep Internal Learning
TLDR
This paper exploits the internal recurrence of information inside a single image, and train a small image-specific CNN at test time, on examples extracted solely from the input image itself, which is the first unsupervised CNN-based SR method. Expand
Toward Real-World Single Image Super-Resolution: A New Benchmark and a New Model
TLDR
This paper builds 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 and presents a Laplacian pyramid based kernel prediction network (LP-KPN), which efficiently learns per-pixel kernels to recover the HR image. Expand
Conditioned Regression Models for Non-blind Single Image Super-Resolution
TLDR
This paper proposes conditioned regression models (including convolutional neural networks and random forests) that can effectively exploit the additional kernel information during both, training and inference and empirically shows that they can effectively handle scenarios where the blur kernel is different for each image. Expand
Learning a Single Convolutional Super-Resolution Network for Multiple Degradations
TLDR
Extensive experimental results show that the proposed convolutional super-resolution network not only can produce favorable results on multiple degradations but also is computationally efficient, providing a highly effective and scalable solution to practical SISR applications. Expand
Nonparametric Blind Super-resolution
  • T. Michaeli, M. Irani
  • Mathematics, Computer Science
  • 2013 IEEE International Conference on Computer Vision
  • 2013
TLDR
It is shown that recurrence of small patches across scales of the low-res image (which forms the basis for single-image SR), can also be used for estimating the optimal blur kernel, which leads to significant improvement in SR results. Expand
Enhanced Deep Residual Networks for Single Image Super-Resolution
TLDR
This paper develops an enhanced deep super-resolution network (EDSR) with performance exceeding those of current state-of-the-art SR methods, and proposes a new multi-scale deepsuper-resolution system (MDSR) and training method, which can reconstruct high-resolution images of different upscaling factors in a single model. Expand
A soft MAP framework for blind super-resolution image reconstruction
TLDR
A new soft maximum a posteriori (MAP) estimation framework to perform joint blur identification and HR image reconstruction that incorporates a soft blur prior that estimates the relevance of the best-fit parametric blur model, and induces reinforcement learning towards it. Expand
A Fully Progressive Approach to Single-Image Super-Resolution
TLDR
To obtain more photorealistic results, a generative adversarial network (GAN) is designed, named ProGanSR, that follows the same progressive multi-scale design principle and constitutes a principled multi- scale approach that increases the reconstruction quality for all upsampling factors simultaneously. Expand
NTIRE 2017 Challenge on Single Image Super-Resolution: Methods and Results
  • R. Timofte, E. Agustsson, +74 authors Q. Guo
  • Computer Science
  • 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)
  • 2017
TLDR
This paper reviews the first challenge on single image super-resolution (restoration of rich details in an low resolution image) with focus on proposed solutions and results and gauges the state-of-the-art in single imagesuper-resolution. Expand
...
1
2
3
4
5
...