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

In the literature, coarse-to-fine or scale-recurrent approach i.e.progressively restoring a clean image from its low-resolution versions has beensuccessfully employed for single image deblurring. However, a majordisadvantage of existing methods is the need for paired data; i.e. sharpblurimage pairs of the same scene, which is a complicated and cumbersomeacquisition procedure. Additionally, due to strong supervision on lossfunctions, pre-trained models of such networks are strongly biased towards theblur experienced during training and tend to give sub-optimal performance whenconfronted by new blur kernels during inference time. To address the aboveissues, we propose unsupervised domain-specific deblurring using ascale-adaptive attention module (SAAM). Our network does not require supervisedpairs for training, and the deblurring mechanism is primarily guided byadversarial loss, thus making our network suitable for a distribution of blurfunctions. Given a blurred input image, different resolutions of the same imageare used in our model during training and SAAM allows for effective flow ofinformation across the resolutions. For network training at a specific scale,SAAM attends to lower scale features as a function of the current scale.Different ablation studies show that our coarse-to-fine mechanism outperformsend-to-end unsupervised models and SAAM is able to attend better compared toattention models used in literature. Qualitative and quantitative comparisons(on no-reference metrics) show that our method outperforms prior unsupervisedmethods.

Unsupervised Domain-Specific Deblurring using Scale-Specific Attention