文献类型：期刊文献

英文题名：CRNet: Channel-Enhanced Remodeling-Based Network for Salient Object Detection in Optical Remote Sensing Images

作者：Sun, Le[1] Wang, Qing[2] Chen, Yuwen[4] Zheng, Yuhui[2,3] Wu, Zebin[2] Fu, Liyong[5] Jeon, Byeungwoo[6]

第一作者：Sun, Le

通信作者：Fu, LY[1]

机构：[1]Nanjing Univ Informat Sci & Technol, Jiangsu Collaborat Innovat Ctr Atmospher Environm, Sch Comp Sci, Nanjing 210044, Peoples R China;[2]Nanjing Univ Informat Sci & Technol, Sch Comp Sci, Nanjing 210044, Peoples R China;[3]Xidian Univ, State Key Lab Integrated Serv Networks, Xian 710071, Peoples R China;[4]Chinese Acad Sci, Chongqing Inst Green & Intelligent Technol, Chongqing 400714, Peoples R China;[5]Chinese Acad Forestry, Inst Forest Resource Informat Tech, Beijing 100091, Peoples R China;[6]Sungkyunkwan Univ, Sch Elect & Elect Engn, Suwon 440746, South Korea

年份：2023

卷号：61

外文期刊名：IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING

收录：;EI(收录号：20233414601977);Scopus(收录号：2-s2.0-85168265934);WOS:【SCI-EXPANDED(收录号:WOS:001065137900002)】；

基金：This work was supported in part by the National Natural Science Foundation of China under Grant 61971233, Grant 62076137, and Grant U20B2065; in part by the Natural Science Foundation of Jiangsu Province under Grant BK20211539; and in part by the Foundation of State Key Laboratory of Integrated Services Networks of Xidian University under Grant ISN22-09.

语种：英文

外文关键词：Channel enhance module (CEM); optical remote sensing images (RSIS); redefined feature module (RFM); salient object detection (SOD)

摘要：Despite the remarkable progress made by the salient object detection of natural sensing images (NSI-SOD), the complex background and scale diversity issues of remote sensing images (RSIs) still pose a substantial obstacle. In this study, we build an end-to-end channel-enhanced remodeling-based network (CRNet) for optical RSIs (ORSIs) to highlight salient objects through feature augmentation. First, the backbone convolutional block is used to suggest the fundamental characteristics. Then, we use the channel enhance module (CEM) to enhance the shallow features. CEM primarily relies on the channel attention (CA) mechanism and uses a no-downscaling strategy to produce local cross-channel interaction, which lowers model complexity while enhancing extraction performance. Meanwhile, we use the redefined feature module (RFM) to reconstruct the deep features and generate global attention features by dimensional transformation and feature relationship aggregation to achieve the role of locating salient targets. Finally, the cascade combines the multiscale features to provide the final saliency map. To further enhance the representational power of the network, we use a hybrid loss function to improve performance. The proposed approach outperforms current state-of-the-art (SOTA) methods, as shown by several experiments on three available datasets. The source code of the proposed CRNet is available publicly at https://github.com/hilitteq/CRNet.git.