AUTHOR=Guan Yang , Liu Xiaoyan , Yu Zhibin , Wang Yubo , Zheng Xingyu , Zhang Shaoda , Zheng Bing 

TITLE=Fast underwater image enhancement based on a generative adversarial framework

JOURNAL=Frontiers in Marine Science

VOLUME=Volume 9 - 2022

YEAR=2023

URL=https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2022.964600

DOI=10.3389/fmars.2022.964600

ISSN=2296-7745

ABSTRACT=Underwater image enhancement is a fundamental requirement in the field of underwater vision. Benefiting from the success of deep learning, underwater image enhancement has made remarkable progress. However, most deep learning-based enhancement methods are computationally expensive, which restricts their application in real-time large-size underwater image processing. Furthermore, GAN-based methods tend to generate spatially inconsistent styles that decrease the enhanced image quality. To solve these problems, we propose a novel efficiency model based on a generative adversarial framework for large-size underwater image enhancement, named FSpiral-GAN. We design our model with equal upsampling blocks (EUBs), equal downsampling blocks (EDBs) and lightweight residual channel attention blocks (RCABs), effectively simplifying the network structure and solving the spatial inconsistency problem. Enhancement experiments on a large number of real underwater datasets demonstrate the advanced performance and improved efficiency of our model.