文献类型：期刊文献

英文题名：Cross-Domain Hyperspectral Image Classification Based on Bi-Directional Domain Adaptation

作者：Zhang, Yuxiang[1] Li, Wei[2,3,4] Jia, Wen[5,6] Zhang, Mengmeng[2,3,4] Tao, Ran[2,3,4] Liang, Shunlin[1]

第一作者：Zhang, Yuxiang

通信作者：Li, W[1]

机构：[1]Univ Hong Kong, Dept Geog, Jockey Club STEM Lab Quantitat Remote Sensing, Hong Kong, Peoples R China;[2]Beijing Inst Technol, Sch Informat & Elect, Beijing 100811, Peoples R China;[3]Natl Key Lab Sci & Technol Space Born Intelligent, Beijing 100081, Peoples R China;[4]Beijing Inst Technol, Zhuhai 519088, Guangdong, Peoples R China;[5]Chinese Acad Forestry, Inst Forest Resource Informat Tech, Beijing 100091, Peoples R China;[6]Natl Forestry & Grassland Adm, Key Lab Forestry Remote Sensing & Informat Syst, Beijing 100091, Peoples R China

年份：2025

卷号：35

期号：12

起止页码：12038-12051

外文期刊名：IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY

收录：;EI(收录号：20252818776350);Scopus(收录号：2-s2.0-105010323129);WOS:【SCI-EXPANDED(收录号:WOS:001631874000007)】；

基金：This work was supported in part by the National Natural Science Foundation of China under Grant 42101403 and Grant 6247012790, in part by Beijing Natural Science Foundation under Grant 4232013, in part by the National Key Research and Development Program of China under Grant 2023YFD2200804 and Grant 2017YFD0600404, and in part by the Jockey Club Science, Technology, Engineering, and Mathematics (STEM) Laboratory of Quantitative Remote Sensing through The Hong Kong Jockey Club Charities Trust.

语种：英文

外文关键词：Hyperspectral image classification; cross-domain; domain adaptation; transformer

摘要：Utilizing hyperspectral remote sensing technology enables the extraction of fine-grained land cover classes. Typically, satellite or airborne images used for training and testing are acquired from different regions or times, where the same class has significant spectral shifts in different scenes. In this paper, we propose a Bi-directional Domain Adaptation (BiDA) framework for cross-domain hyperspectral image (HSI) classification, which focuses on extracting both domain-invariant features and domain-specific information in the independent adaptive space, thereby enhancing the adaptability and separability to the target scene. In the proposed BiDA, a triple-branch transformer architecture (the source branch, target branch, and coupled branch) with semantic tokenizer is designed as the backbone. Specifically, the source branch and target branch independently learn the adaptive space of source and target domains, a Coupled Multi-head Cross-attention (CMCA) mechanism is developed in coupled branch for feature interaction and inter-domain correlation mining. Furthermore, a bi-directional distillation loss is designed to guide adaptive space learning using inter-domain correlation. Finally, we propose an Adaptive Reinforcement Strategy (ARS) to encourage the model to focus on specific generalized feature extraction within both source and target scenes in noise condition. Experimental results on cross-temporal/scene airborne and satellite datasets demonstrate that the proposed BiDA performs significantly better than some state-of-the-art domain adaptation approaches. In the cross-temporal tree species classification task, the proposed BiDA is more than 3% similar to 5% higher than the most advanced method. The codes will be available from the website: https://github.com/YuxiangZhang-BIT/IEEE_TCSVT_BiDA