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

Recovering clear images from blurry ones with an unknown blur kernel is achallenging problem. Deep image prior (DIP) proposes to use the deep network asa regularizer for a single image rather than as a supervised model, whichachieves encouraging results in the nonblind deblurring problem. However, sincethe relationship between images and the network architectures is unclear, it ishard to find a suitable architecture to provide sufficient constraints on theestimated blur kernels and clean images. Also, DIP uses the sparse maximum aposteriori (MAP), which is insufficient to enforce the selection of therecovery image. Recently, variational deep image prior (VDIP) was proposed toimpose constraints on both blur kernels and recovery images and take thestandard deviation of the image into account during the optimization process bythe variational principle. However, we empirically find that VDIP struggleswith processing image details and tends to generate suboptimal results when theblur kernel is large. Therefore, we combine total generalized variational (TGV)regularization with VDIP in this paper to overcome these shortcomings of VDIP.TGV is a flexible regularization that utilizes the characteristics of partialderivatives of varying orders to regularize images at different scales,reducing oil painting artifacts while maintaining sharp edges. The proposedVDIP-TGV effectively recovers image edges and details by supplementing extragradient information through TGV. Additionally, this model is solved by thealternating direction method of multipliers (ADMM), which effectively combinestraditional algorithms and deep learning methods. Experiments show that ourproposed VDIP-TGV surpasses various state-of-the-art models quantitatively andqualitatively.

VDIP-TGV: Blind Image Deconvolution via Variational Deep Image Prior Empowered by Total Generalized Variation