English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Recent deep-learning based Super-Resolution (SR) methods have achievedremarkable performance on images with known degradation. However, these methodsalways fail in real-world scene, since the Low-Resolution (LR) images after theideal degradation (e.g., bicubic down-sampling) deviate from real sourcedomain. The domain gap between the LR images and the real-world images can beobserved clearly on frequency density, which inspires us to explictly narrowthe undesired gap caused by incorrect degradation. From this point of view, wedesign a novel Frequency Consistent Adaptation (FCA) that ensures the frequencydomain consistency when applying existing SR methods to the real scene. Weestimate degradation kernels from unsupervised images and generate thecorresponding LR images. To provide useful gradient information for kernelestimation, we propose Frequency Density Comparator (FDC) by distinguishing thefrequency density of images on different scales. Based on the domain-consistentLR-HR pairs, we train easy-implemented Convolutional Neural Network (CNN) SRmodels. Extensive experiments show that the proposed FCA improves theperformance of the SR model under real-world setting achieving state-of-the-artresults with high fidelity and plausible perception, thus providing a noveleffective framework for real-world SR application.

Frequency Consistent Adaptation for Real World Super Resolution