Next Article in Journal
Investigation of Antarctic Precipitable Water Vapor Variability and Trend from 18 Year (2001 to 2018) Data of Four Reanalyses Based on Radiosonde and GNSS Observations
Next Article in Special Issue
Hyperspectral Image Super-Resolution Based on Spatial Correlation-Regularized Unmixing Convolutional Neural Network
Previous Article in Journal
Regionalized Linear Models for River Depth Retrieval Using 3-Band Multispectral Imagery and Green LIDAR Data
Previous Article in Special Issue
Multi-Object Segmentation in Complex Urban Scenes from High-Resolution Remote Sensing Data
Article

DEANet: Dual Encoder with Attention Network for Semantic Segmentation of Remote Sensing Imagery

by 1,†, 1,*, 1,†, 2 and 1
1
State Key Laboratory of Intelligent Control and Decision of Complex Systems, Beijing Institute of Technology, Beijing 100081, China
2
School of Information and Electronics, Beijing Institute of Technology, Beijing 100081, China
*
Author to whom correspondence should be addressed.
These authors contributed equally to this work.
Academic Editors: Angelica I. Aviles-Rivero, Weijia Li, Lichao Mou, Runmin Dong and Juepeng Zheng
Remote Sens. 2021, 13(19), 3900; https://0-doi-org.brum.beds.ac.uk/10.3390/rs13193900
Received: 30 August 2021 / Revised: 25 September 2021 / Accepted: 26 September 2021 / Published: 29 September 2021
(This article belongs to the Special Issue Deep Learning in Remote Sensing Application)
Remote sensing has now been widely used in various fields, and the research on the automatic land-cover segmentation methods of remote sensing imagery is significant to the development of remote sensing technology. Deep learning methods, which are developing rapidly in the field of semantic segmentation, have been widely applied to remote sensing imagery segmentation. In this work, a novel deep learning network—Dual Encoder with Attention Network (DEANet) is proposed. In this network, a dual-branch encoder structure, whose first branch is used to generate a rough guidance feature map as area attention to help re-encode feature maps in the next branch, is proposed to improve the encoding ability of the network, and an improved pyramid partial decoder (PPD) based on the parallel partial decoder is put forward to make fuller use of the features form the encoder along with the receptive filed block (RFB). In addition, an edge attention module using the transfer learning method is introduced to explicitly advance the segmentation performance in edge areas. Except for structure, a loss function composed with the weighted Cross Entropy (CE) loss and weighted Union subtract Intersection (UsI) loss is designed for training, where UsI loss represents a new region-based aware loss which replaces the IoU loss to adapt to multi-classification tasks. Furthermore, a detailed training strategy for the network is introduced as well. Extensive experiments on three public datasets verify the effectiveness of each proposed module in our framework and demonstrate that our method achieves more excellent performance over some state-of-the-art methods. View Full-Text
Keywords: remote sensing; land cover classification; deep learning; semantic segmentation; encoder-decoder; attention mechanism remote sensing; land cover classification; deep learning; semantic segmentation; encoder-decoder; attention mechanism
Show Figures

Graphical abstract

MDPI and ACS Style

Wei, H.; Xu, X.; Ou, N.; Zhang, X.; Dai, Y. DEANet: Dual Encoder with Attention Network for Semantic Segmentation of Remote Sensing Imagery. Remote Sens. 2021, 13, 3900. https://0-doi-org.brum.beds.ac.uk/10.3390/rs13193900

AMA Style

Wei H, Xu X, Ou N, Zhang X, Dai Y. DEANet: Dual Encoder with Attention Network for Semantic Segmentation of Remote Sensing Imagery. Remote Sensing. 2021; 13(19):3900. https://0-doi-org.brum.beds.ac.uk/10.3390/rs13193900

Chicago/Turabian Style

Wei, Haoran, Xiangyang Xu, Ni Ou, Xinru Zhang, and Yaping Dai. 2021. "DEANet: Dual Encoder with Attention Network for Semantic Segmentation of Remote Sensing Imagery" Remote Sensing 13, no. 19: 3900. https://0-doi-org.brum.beds.ac.uk/10.3390/rs13193900

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop