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

We propose an enhanced multi-scale network, dubbed GridDehazeNet+, for singleimage dehazing. The proposed dehazing method does not rely on the AtmosphereScattering Model (ASM), and an explanation as to why it is not necessarilyperforming the dimension reduction offered by this model is provided.GridDehazeNet+ consists of three modules: pre-processing, backbone, andpost-processing. The trainable pre-processing module can generate learnedinputs with better diversity and more pertinent features as compared to thosederived inputs produced by hand-selected pre-processing methods. The backbonemodule implements multi-scale estimation with two major enhancements: 1) anovel grid structure that effectively alleviates the bottleneck issue via denseconnections across different scales; 2) a spatial-channel attention block thatcan facilitate adaptive fusion by consolidating dehazing-relevant features. Thepost-processing module helps to reduce the artifacts in the final output. Dueto domain shift, the model trained on synthetic data may not generalize well onreal data. To address this issue, we shape the distribution of synthetic datato match that of real data, and use the resulting translated data to finetuneour network. We also propose a novel intra-task knowledge transfer mechanismthat can memorize and take advantage of synthetic domain knowledge to assistthe learning process on the translated data. Experimental results demonstratethat the proposed method outperforms the state-of-the-art on several syntheticdehazing datasets, and achieves the superior performance on real-world hazyimages after finetuning.

GridDehazeNet+: An Enhanced Multi-Scale Network with Intra-Task Knowledge Transfer for Single Image Dehazing