文献类型：期刊文献

英文题名：FSAD-Net: Feedback Spatial Attention Dehazing Network

作者：Zhou, Yu[1];Chen, Zhihua[1];Li, Ping[2];Song, Haitao[3];Chen, C. L. Philip[4,5,6];Sheng, Bin[7]

机构：[1]East China Univ Sci & Technol, Dept Comp Sci & Engn, Shanghai 200237, Peoples R China;[2]Hong Kong Polytech Univ, Dept Comp, Hong Kong, Peoples R China;[3]Shanghai Jiao Tong Univ, Artificial Intelligence Inst, Shanghai 200240, Peoples R China;[4]South China Univ Technol, Sch Comp Sci & Engn, Guangzhou 510006, Peoples R China;[5]Dalian Maritime Univ, Nav Coll, Dalian 116026, Peoples R China;[6]Univ Macau, Fac Sci & Technol, Macau, Peoples R China;[7]Shanghai Jiao Tong Univ, Dept Comp Sci & Engn, Shanghai 200240, Peoples R China

年份：2023

卷号：34

期号：10

起止页码：7719

外文期刊名：IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS

收录：;EI(收录号：20220811677014);WOS:【SCI-EXPANDED(收录号:WOS:000754282000001)】；

基金：This work was supported in part by the National Natural Science Foundation of China under Grant 61672228, Grant 61872241, and Grant 61572316; in part by the National Key Research and Development Program of China under Grant 2019YFB1703600; and in part by The Hong Kong Polytechnic University under Grant P0030419, Grant P0030929, and Grant P0035358.

语种：英文

外文关键词：Atmospheric modeling; Image color analysis; Image restoration; Scattering; Feature extraction; Correlation; Indexes; Dehazing network; image dehazing; recurrent structure; spatial attention mechanism

摘要：Recent dehazing networks learn more discriminative high-level features by designing deeper networks or introducing complicated structures, while ignoring inherent feature correlations in intermediate layers. In this article, we establish a novel and effective end-to-end dehazing method, named feedback spatial attention dehazing network (FSAD-Net). FSAD-Net is based on the recurrent structure and consists of four modules: a shallow feature extraction block (SFEB), a feedback block (FB), multiple advanced residual blocks (ARBs), and a reconstruction block (RB). FB is designed to handle feedback connections, and it can improve the dehazing performance by exploiting the dependencies of deep features across stages. ARB implements a novel attention-based estimation on a residual block to adapt to pixels with different distributions. Finally, RB helps restore haze-free images. It can be seen from the experimental results that FSAD-Net almost outperforms the state-of-the-arts in terms of five quantitative metrics. Moreover, the qualitatively comparisons on real-world images also demonstrate the superiority of the proposed FSAD-Net. Considering the efficiency and effectiveness of FSAD-Net, it can be expected to serve as a suitable image dehazing baseline in the future.