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

Conventional CNNs-based dehazing models suffer from two essential issues: thedehazing framework (limited in interpretability) and the convolution layers(content-independent and ineffective to learn long-range dependencyinformation). In this paper, firstly, we propose a new complementary featureenhanced framework, in which the complementary features are learned by severalcomplementary subtasks and then together serve to boost the performance of theprimary task. One of the prominent advantages of the new framework is that thepurposively chosen complementary tasks can focus on learning weakly dependentcomplementary features, avoiding repetitive and ineffective learning of thenetworks. We design a new dehazing network based on such a framework.Specifically, we select the intrinsic image decomposition as the complementarytasks, where the reflectance and shading prediction subtasks are used toextract the color-wise and texture-wise complementary features. To effectivelyaggregate these complementary features, we propose a complementary featuresselection module (CFSM) to select the more useful features for image dehazing.Furthermore, we introduce a new version of vision transformer block, namedHybrid Local-Global Vision Transformer (HyLoG-ViT), and incorporate it withinour dehazing networks. The HyLoG-ViT block consists of the local and the globalvision transformer paths used to capture local and global dependencies. As aresult, the HyLoG-ViT introduces locality in the networks and captures theglobal and long-range dependencies. Extensive experiments on homogeneous,non-homogeneous, and nighttime dehazing tasks reveal that the proposed dehazingnetwork can achieve comparable or even better performance than CNNs-baseddehazing models.

Complementary Feature Enhanced Network with Vision Transformer for Image Dehazing