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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Rain in the dark is a common natural phenomenon. Photos captured in such acondition significantly impact the performance of various nighttime activities,such as autonomous driving, surveillance systems, and night photography. Whileexisting methods designed for low-light enhancement or deraining show promisingperformance, they have limitations in simultaneously addressing the task ofbrightening low light and removing rain. Furthermore, using a cascade approach,such as ``deraining followed by low-light enhancement'' or vice versa, may leadto difficult-to-handle rain patterns or excessively blurred and overexposedimages. To overcome these limitations, we propose an end-to-end network called$L^{2}RIRNet$ which can jointly handle low-light enhancement and deraining. Ournetwork mainly includes a Pairwise Degradation Feature Vector ExtractionNetwork (P-Net) and a Restoration Network (R-Net). P-Net can learn degradationfeature vectors on the dark and light areas separately, using contrastivelearning to guide the image restoration process. The R-Net is responsible forrestoring the image. We also introduce an effective Fast Fourier - ResNetDetail Guidance Module (FFR-DG) that initially guides image restoration usingdetail image that do not contain degradation information but focus on texturedetail information. Additionally, we contribute a dataset containing syntheticand real-world low-light-rainy images. Extensive experiments demonstrate thatour $L^{2}RIRNet$ outperforms existing methods in both synthetic and complexreal-world scenarios.

Unlocking Low-Light-Rainy Image Restoration by Pairwise Degradation Feature Vector Guidance