Fast Adaptation to Super-Resolution Networks via Meta-learning
Seobin Park,Jinsu Yoo,Donghyeon Cho,Jiwon Kim,Tae Hyun Kim +4 more
- pp 754-769
TLDR
The proposed model-agnostic approach consistently improves the performance of conventional SR networks on various benchmark SR datasets and effectively handles unknown SR kernels and can be applied to any existing model.Abstract:
Conventional supervised super-resolution (SR) approaches are trained with massive external SR datasets but fail to exploit desirable properties of the given test image. On the other hand, self-supervised SR approaches utilize the internal information within a test image but suffer from computational complexity in run-time. In this work, we observe the opportunity for further improvement of the performance of single-image super-resolution (SISR) without changing the architecture of conventional SR networks by practically exploiting additional information given from the input image. In the training stage, we train the network via meta-learning; thus, the network can quickly adapt to any input image at test time. Then, in the test stage, parameters of this meta-learned network are rapidly fine-tuned with only a few iterations by only using the given low-resolution image. The adaptation at the test time takes full advantage of patch-recurrence property observed in natural images. Our method effectively handles unknown SR kernels and can be applied to any existing model. We demonstrate that the proposed model-agnostic approach consistently improves the performance of conventional SR networks on various benchmark SR datasets.read more
Citations
More filters
Journal ArticleDOI
Real-world single image super-resolution: A brief review
TL;DR: Li et al. as discussed by the authors made a comprehensive review on real-world single image super-resolution (RSISR), and four major categories of RSISR methods, namely the degradation modeling-based, image pairsbased, domain translation-based and self-learning-based SR methods.
Journal ArticleDOI
Real-World Single Image Super-Resolution: A Brief Review
Honggang Chen,Исмаилова М.Б.,Xiaohai He,Linbo Qing,Yuanyuan Wu,Chao Ren,Ray E. Sheriff,Ce Zhu +7 more
TL;DR: Li et al. as discussed by the authors made a comprehensive review on real-world single image super-resolution (RSISR), and four major categories of RSISR methods, namely the degradation modeling-based, image pairsbased, domain translation-based and self-learning-based SR methods.
Proceedings ArticleDOI
Test-Time Fast Adaptation for Dynamic Scene Deblurring via Meta-Auxiliary Learning
TL;DR: Li et al. as mentioned in this paper proposed a self-supervised meta-auxiliary learning to improve the performance of deblurring by integrating both external and internal learning, which is able to exploit the internal information at test time via the auxiliary task to enhance the performance.
Proceedings ArticleDOI
Tackling the Ill-Posedness of Super-Resolution through Adaptive Target Generation
TL;DR: In this article, the adaptive target is generated from the original ground truth (GT) target by a transformation to match the output of the SR network, which provides the algorithm with the flexibility of accepting a variety of valid solutions.
Journal ArticleDOI
Conditional Hyper-Network for Blind Super-Resolution With Multiple Degradations
TL;DR: Guan et al. as mentioned in this paper proposed a novel conditional hyper-network framework for super-resolution with multiple degradatiions (named CMDSR), which helps the SR framework learn how to adapt to changes in the degradation distribution of input.
References
More filters
Proceedings ArticleDOI
Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks
TL;DR: CycleGAN as discussed by the authors learns a mapping G : X → Y such that the distribution of images from G(X) is indistinguishable from the distribution Y using an adversarial loss.
Proceedings Article
Model-agnostic meta-learning for fast adaptation of deep networks
TL;DR: An algorithm for meta-learning that is model-agnostic, in the sense that it is compatible with any model trained with gradient descent and applicable to a variety of different learning problems, including classification, regression, and reinforcement learning is proposed.
Proceedings ArticleDOI
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network
Christian Ledig,Lucas Theis,Ferenc Huszar,Jose Caballero,Andrew Cunningham,Alejandro Acosta,Andrew Peter Aitken,Alykhan Tejani,Johannes Totz,Zehan Wang,Wenzhe Shi +10 more
TL;DR: SRGAN as mentioned in this paper proposes a perceptual loss function which consists of an adversarial loss and a content loss, which pushes the solution to the natural image manifold using a discriminator network that is trained to differentiate between the super-resolved images and original photo-realistic images.
Book ChapterDOI
Perceptual Losses for Real-Time Style Transfer and Super-Resolution
TL;DR: In this paper, the authors combine the benefits of both approaches, and propose the use of perceptual loss functions for training feed-forward networks for image style transfer, where a feedforward network is trained to solve the optimization problem proposed by Gatys et al. in real-time.
Proceedings ArticleDOI
A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics
TL;DR: In this paper, the authors present a database containing ground truth segmentations produced by humans for images of a wide variety of natural scenes, and define an error measure which quantifies the consistency between segmentations of differing granularities.