TY - JOUR
T1 - Hybrid Feature Aligned Network for Salient Object Detection in Optical Remote Sensing Imagery
AU - Wang, Qi
AU - Liu, Yanfeng
AU - Xiong, Zhitong
AU - Yuan, Yuan
N1 - Publisher Copyright:
© 1980-2012 IEEE.
PY - 2022
Y1 - 2022
N2 - Recently, salient object detection in optical remote sensing images (RSI-SOD) has attracted great attention. Benefiting from the success of deep learning and the inspiration of natural SOD task, RSI-SOD has achieved fast progress over the past two years. However, existing methods usually suffer from the intrinsic problems of optical RSIs: 1) cluttered background; 2) scale variation of salient objects; 3) complicated edges and irregular topology. To remedy these problems, we propose a hybrid feature aligned network (HFANet) jointly modeling boundary learning to detect salient objects effectively. Specifically, we design a hybrid encoder by unifying two components to capture global context for mitigating the disturbance of complex background. Then, to detect multiscale salient objects effectively, we propose a Gated Fold-ASPP (GF-ASPP) to extract abundant context in the deep semantic features. Furthermore, an adjacent feature aligned module (AFAM) is presented for integrating adjacent features with unparameterized alignment strategy. Finally, we propose a novel interactive guidance loss (IGLoss) to combine saliency and edge detection, which can adaptively perform mutual supervision of the two subtasks to facilitate detection of salient objects with blurred edges and irregular topology. Adequate experimental results on three optical RSI-SOD datasets reveal that the presented approach exceeds 18 state-of-the-art ones. All codes and detection results are available at https://github.com/lyf0801/HFANet.
AB - Recently, salient object detection in optical remote sensing images (RSI-SOD) has attracted great attention. Benefiting from the success of deep learning and the inspiration of natural SOD task, RSI-SOD has achieved fast progress over the past two years. However, existing methods usually suffer from the intrinsic problems of optical RSIs: 1) cluttered background; 2) scale variation of salient objects; 3) complicated edges and irregular topology. To remedy these problems, we propose a hybrid feature aligned network (HFANet) jointly modeling boundary learning to detect salient objects effectively. Specifically, we design a hybrid encoder by unifying two components to capture global context for mitigating the disturbance of complex background. Then, to detect multiscale salient objects effectively, we propose a Gated Fold-ASPP (GF-ASPP) to extract abundant context in the deep semantic features. Furthermore, an adjacent feature aligned module (AFAM) is presented for integrating adjacent features with unparameterized alignment strategy. Finally, we propose a novel interactive guidance loss (IGLoss) to combine saliency and edge detection, which can adaptively perform mutual supervision of the two subtasks to facilitate detection of salient objects with blurred edges and irregular topology. Adequate experimental results on three optical RSI-SOD datasets reveal that the presented approach exceeds 18 state-of-the-art ones. All codes and detection results are available at https://github.com/lyf0801/HFANet.
KW - Feature alignment
KW - jointing boundary learning
KW - multiscale context modeling
KW - optical remote sensing image (RSI)
KW - salient object detection (SOD)
UR - http://www.scopus.com/inward/record.url?scp=85131745663&partnerID=8YFLogxK
U2 - 10.1109/TGRS.2022.3181062
DO - 10.1109/TGRS.2022.3181062
M3 - 文章
AN - SCOPUS:85131745663
SN - 0196-2892
VL - 60
JO - IEEE Transactions on Geoscience and Remote Sensing
JF - IEEE Transactions on Geoscience and Remote Sensing
M1 - 5624915
ER -