Decomposition-Based Unsupervised Domain Adaptation for Remote Sensing Image Semantic Segmentation

被引:0
作者
Ma, Xianping [1 ]
Zhang, Xiaokang [2 ,3 ]
Ding, Xingchen [4 ]
Pun, Man-On [1 ]
Ma, Siwei [5 ]
机构
[1] Chinese Univ Hong Kong Shenzhen, Sch Sci & Engn, Shenzhen 518172, Peoples R China
[2] Hubei Luojia Lab, Wuhan 430079, Peoples R China
[3] Wuhan Univ Sci & Technol, Sch Informat Sci & Engn, Wuhan 430081, Peoples R China
[4] Wuhan Univ, Sch Cyber Sci & Engn, Wuhan 430072, Peoples R China
[5] Peking Univ, Sch Elect Engn & Comp Sci, Beijing 100871, Peoples R China
来源
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING | 2024年 / 62卷
基金
中国国家自然科学基金;
关键词
Semantic segmentation; Remote sensing; Training; Semantics; Feature extraction; Context modeling; Transformers; Generators; Generative adversarial networks; Adaptation models; Global-local information; high/low-frequency decomposition (HLFD); remote sensing; semantic segmentation; unsupervised domain adaptation (UDA); NETWORK;
D O I
10.1109/TGRS.2024.3483283
中图分类号
P3 [地球物理学]; P59 [地球化学];
学科分类号
0708 ; 070902 ;
摘要
Unsupervised domain adaptation (UDA) techniques are vital for semantic segmentation in geosciences, effectively utilizing remote sensing imagery across diverse domains. However, most existing UDA methods, which focus on domain alignment at the high-level feature space, struggle to simultaneously retain local spatial details and global contextual semantics. To overcome these challenges, a novel decomposition scheme is proposed to guide domain-invariant representation learning. Specifically, multiscale high/low-frequency decomposition (HLFD) modules are proposed to decompose feature maps into high- and low-frequency components across different subspaces. This decomposition is integrated into a fully global-local generative adversarial network (GLGAN) that incorporates global-local transformer blocks (GLTBs) to enhance the alignment of decomposed features. By integrating the HLFD scheme and the GLGAN, a novel decomposition-based UDA framework called De-GLGAN is developed to improve the cross-domain transferability and generalization capability of semantic segmentation models. Extensive experiments on two UDA benchmarks, namely ISPRS Potsdam and Vaihingen, and LoveDA Rural and Urban, demonstrate the effectiveness and superiority of the proposed approach over existing state-of-the-art UDA methods. The source code for this work is accessible at https://github.com/sstary/SSRS.
引用
收藏
页数:18
相关论文
共 50 条
  • [41] Deep Covariance Alignment for Domain Adaptive Remote Sensing Image Segmentation
    Wu, Linshan
    Lu, Ming
    Fang, Leyuan
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2022, 60
  • [42] FTransDeepLab: Multimodal Fusion Transformer-Based DeepLabv3+for Remote Sensing Semantic Segmentation
    Feng, Haixia
    Hu, Qingwu
    Zhao, Pengcheng
    Wang, Shunli
    Ai, Mingyao
    Zheng, Daoyuan
    Liu, Tiancheng
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2025, 63
  • [43] UGCNet: An Unsupervised Semantic Segmentation Network Embedded With Geometry Consistency for Remote-Sensing Images
    Zhao, Danpei
    Yuan, Bo
    Gao, Yue
    Qi, Xinhu
    Shi, Zhenwei
    IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2022, 19
  • [44] Adaptive Multitype Contrastive Views Generation for Remote Sensing Image Semantic Segmentation
    Shi, Cheng
    Han, Peiwen
    Zhao, Minghua
    Fang, Li
    Miao, Qiguang
    Pun, Chi-Man
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2025, 63
  • [45] ColorMapGAN: Unsupervised Domain Adaptation for Semantic Segmentation Using Color Mapping Generative Adversarial Networks
    Tasar, Onur
    Happy, S. L.
    Tarabalka, Yuliya
    Alliez, Pierre
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2020, 58 (10): : 7178 - 7193
  • [46] Semantic Segmentation of Remote Sensing Image Based on Regional Self-Attention Mechanism
    Zhao, Danpei
    Wang, Chenxu
    Gao, Yue
    Shi, Zhenwei
    Xie, Fengying
    IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2022, 19
  • [47] Memory-Contrastive Unsupervised Domain Adaptation for Building Extraction of High-Resolution Remote Sensing Imagery
    Chen, Jie
    He, Peien
    Zhu, Jingru
    Guo, Ya
    Sun, Geng
    Deng, Min
    Li, Haifeng
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2023, 61
  • [48] Unsupervised domain adaptation alignment method for cross-domain semantic segmentation of remote sensing images
    Shen Z.
    Ni H.
    Guan H.
    Cehui Xuebao/Acta Geodaetica et Cartographica Sinica, 2023, 52 (12): : 1 - 2
  • [49] Domain Adaptive Semantic Segmentation of Remote Sensing Images via Self-Training-Based Dual-Level Data Augmentation
    Hu, Xiaoxing
    Wang, Yupei
    Chen, Liang
    IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2024, 17 : 19713 - 19729
  • [50] Enhanced Swin Transformer and Edge Spatial Attention for Remote Sensing Image Semantic Segmentation
    Liu, Fuxiang
    Hu, Zhiqiang
    Li, Lei
    Li, Hanlu
    Liu, Xinxin
    IEEE SIGNAL PROCESSING LETTERS, 2025, 32 : 1296 - 1300