Journal
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
Volume 14, Issue -, Pages 4530-4546Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JSTARS.2021.3071353
Keywords
Semantics; Image segmentation; Feature extraction; Decoding; Remote sensing; Spatial resolution; Logic gates; Attention module (AM); gate module (GM); high-resolution (HR) remote sensing imagery; semantic segmentation
Categories
Funding
- National Science and Technology Major Project of China [21-Y20A06-9001-17/18]
- National Natural Science Foundation of China [42071297, 41871235, 41871326]
- Fundamental Research Funds for the Central Universities [020914380080]
- Highlevel Innovation and Entrepreneurship Talents Introduction Program of Jiangsu Province of China
Ask authors/readers for more resources
Researchers proposed an end-to-end DCNN network named GAMNet to address issues in semantic segmentation of high-resolution remote sensing images, such as the loss of spatial information, class imbalance, and high diversity of geographic objects. The integration of attention and gate module enables multiscale feature extraction and boundary recovery, leading to state-of-the-art performance on the ISPRS 2-D semantic labeling datasets.
Semantic segmentation of high-resolution (HR) remote sensing images achieved great progress by utilizing deep convolutional neural networks (DCNNs) in recent years. However, the decrease of resolution in the feature map of DCNNs brings about the loss of spatial information and thus leads to the blurring of object boundary and misclassification of small objects. In addition, the class imbalance and the high diversity of geographic objects in HR images exacerbate the performance. To deal with the above problems, we proposed an end-to-end DCNN network named GAMNet to balance the contradiction between global semantic information and local details. An integration of attention and gate module (GAM) is specially designed to simultaneously realize multiscale feature extraction and boundary recovery. The integration module can be inserted in an encoder-decoder network with skip connection. Meanwhile, a composite loss function is designed to achieve deep supervision of GAM by adding an auxiliary loss, which can help improve the effectiveness of the integration module. The performance of GAMNet is quantitatively evaluated on the ISPRS 2-D semantic labeling datasets and achieves state-of-the-art performance in comparison with other representative methods.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available