Class-Discrepancy Dynamic Weighting for Cross-Domain Few-Shot Hyperspectral Image Classification
In recent years, cross-domain few-shot learning (CDFSL) has demonstrated remarkable performance in hyperspectral image classification (HSIC), partially alleviating the distribution shift problem. However, most domain adaptation methods rely on similarity metrics to establish cross-domain class match...
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-07-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/17/15/2605 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849239669915516928 |
|---|---|
| author | Chen Ding Jiahao Yue Sirui Zheng Yizhuo Dong Wenqiang Hua Xueling Chen Yu Xie Song Yan Wei Wei Lei Zhang |
| author_facet | Chen Ding Jiahao Yue Sirui Zheng Yizhuo Dong Wenqiang Hua Xueling Chen Yu Xie Song Yan Wei Wei Lei Zhang |
| author_sort | Chen Ding |
| collection | DOAJ |
| description | In recent years, cross-domain few-shot learning (CDFSL) has demonstrated remarkable performance in hyperspectral image classification (HSIC), partially alleviating the distribution shift problem. However, most domain adaptation methods rely on similarity metrics to establish cross-domain class matching, making it difficult to simultaneously account for intra-class sample size variations and inherent inter-class differences. To address this problem, existing studies have introduced a class weighting mechanism within the prototype network framework, determining class weights by calculating inter-sample similarity through distance metrics. However, this method suffers from a dual limitation: susceptibility to noise interference and insufficient capacity to capture global class variations, which may lead to distorted weight allocation and consequently result in alignment bias. To solve these issues, we propose a novel class-discrepancy dynamic weighting-based cross-domain FSL (CDDW-CFSL) framework. It integrates three key components: (1) the class-weighted domain adaptation (CWDA) method dynamically measures cross-domain distribution shifts using global class mean discrepancies. It employs discrepancy-sensitive weighting to strengthen the alignment of critical categories, enabling accurate domain adaptation while maintaining feature topology; (2) the class mean refinement (CMR) method incorporates class covariance distance to compute distribution discrepancies between support set samples and class prototypes, enabling the precise capture of cross-domain feature internal structures; (3) a novel multi-dimensional feature extractor that captures both local spatial details and continuous spectral characteristics simultaneously, facilitating deep cross-dimensional feature fusion. The results in three publicly available HSIC datasets show the effectiveness of the CDDW-CFSL. |
| format | Article |
| id | doaj-art-4ffae5384c664a7b8f3366cb355b5c85 |
| institution | Kabale University |
| issn | 2072-4292 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj-art-4ffae5384c664a7b8f3366cb355b5c852025-08-20T04:00:53ZengMDPI AGRemote Sensing2072-42922025-07-011715260510.3390/rs17152605Class-Discrepancy Dynamic Weighting for Cross-Domain Few-Shot Hyperspectral Image ClassificationChen Ding0Jiahao Yue1Sirui Zheng2Yizhuo Dong3Wenqiang Hua4Xueling Chen5Yu Xie6Song Yan7Wei Wei8Lei Zhang9School of Computer Science and Technology, Xi’an University of Posts and Telecommunications, Xi’an 710129, ChinaSchool of Computer Science and Technology, Xi’an University of Posts and Telecommunications, Xi’an 710129, ChinaSchool of Computer Science and Technology, Xi’an University of Posts and Telecommunications, Xi’an 710129, ChinaSchool of Computer Science and Technology, Xi’an University of Posts and Telecommunications, Xi’an 710129, ChinaSchool of Computer Science and Technology, Xi’an University of Posts and Telecommunications, Xi’an 710129, ChinaSchool of Computer Science and Technology, Xi’an University of Posts and Telecommunications, Xi’an 710129, ChinaSchool of Information and Electronics, Beijing Institute of Technology, Beijing 100081, ChinaXi’an Aerospace Propulsion Institute, Xi’an 710100, ChinaShaanxi Key Lab of Speech and Image Information Processing (SAIIP), School of Computer Science and Engineering, Northwestern Polytechnical University, Xi’an 710129, ChinaShaanxi Key Lab of Speech and Image Information Processing (SAIIP), School of Computer Science and Engineering, Northwestern Polytechnical University, Xi’an 710129, ChinaIn recent years, cross-domain few-shot learning (CDFSL) has demonstrated remarkable performance in hyperspectral image classification (HSIC), partially alleviating the distribution shift problem. However, most domain adaptation methods rely on similarity metrics to establish cross-domain class matching, making it difficult to simultaneously account for intra-class sample size variations and inherent inter-class differences. To address this problem, existing studies have introduced a class weighting mechanism within the prototype network framework, determining class weights by calculating inter-sample similarity through distance metrics. However, this method suffers from a dual limitation: susceptibility to noise interference and insufficient capacity to capture global class variations, which may lead to distorted weight allocation and consequently result in alignment bias. To solve these issues, we propose a novel class-discrepancy dynamic weighting-based cross-domain FSL (CDDW-CFSL) framework. It integrates three key components: (1) the class-weighted domain adaptation (CWDA) method dynamically measures cross-domain distribution shifts using global class mean discrepancies. It employs discrepancy-sensitive weighting to strengthen the alignment of critical categories, enabling accurate domain adaptation while maintaining feature topology; (2) the class mean refinement (CMR) method incorporates class covariance distance to compute distribution discrepancies between support set samples and class prototypes, enabling the precise capture of cross-domain feature internal structures; (3) a novel multi-dimensional feature extractor that captures both local spatial details and continuous spectral characteristics simultaneously, facilitating deep cross-dimensional feature fusion. The results in three publicly available HSIC datasets show the effectiveness of the CDDW-CFSL.https://www.mdpi.com/2072-4292/17/15/2605hyperspectral image classificationdomain adaptationfew-shot learning (FSL)class-discrepancy dynamic weightingmulti-dimensional feature extraction |
| spellingShingle | Chen Ding Jiahao Yue Sirui Zheng Yizhuo Dong Wenqiang Hua Xueling Chen Yu Xie Song Yan Wei Wei Lei Zhang Class-Discrepancy Dynamic Weighting for Cross-Domain Few-Shot Hyperspectral Image Classification Remote Sensing hyperspectral image classification domain adaptation few-shot learning (FSL) class-discrepancy dynamic weighting multi-dimensional feature extraction |
| title | Class-Discrepancy Dynamic Weighting for Cross-Domain Few-Shot Hyperspectral Image Classification |
| title_full | Class-Discrepancy Dynamic Weighting for Cross-Domain Few-Shot Hyperspectral Image Classification |
| title_fullStr | Class-Discrepancy Dynamic Weighting for Cross-Domain Few-Shot Hyperspectral Image Classification |
| title_full_unstemmed | Class-Discrepancy Dynamic Weighting for Cross-Domain Few-Shot Hyperspectral Image Classification |
| title_short | Class-Discrepancy Dynamic Weighting for Cross-Domain Few-Shot Hyperspectral Image Classification |
| title_sort | class discrepancy dynamic weighting for cross domain few shot hyperspectral image classification |
| topic | hyperspectral image classification domain adaptation few-shot learning (FSL) class-discrepancy dynamic weighting multi-dimensional feature extraction |
| url | https://www.mdpi.com/2072-4292/17/15/2605 |
| work_keys_str_mv | AT chending classdiscrepancydynamicweightingforcrossdomainfewshothyperspectralimageclassification AT jiahaoyue classdiscrepancydynamicweightingforcrossdomainfewshothyperspectralimageclassification AT siruizheng classdiscrepancydynamicweightingforcrossdomainfewshothyperspectralimageclassification AT yizhuodong classdiscrepancydynamicweightingforcrossdomainfewshothyperspectralimageclassification AT wenqianghua classdiscrepancydynamicweightingforcrossdomainfewshothyperspectralimageclassification AT xuelingchen classdiscrepancydynamicweightingforcrossdomainfewshothyperspectralimageclassification AT yuxie classdiscrepancydynamicweightingforcrossdomainfewshothyperspectralimageclassification AT songyan classdiscrepancydynamicweightingforcrossdomainfewshothyperspectralimageclassification AT weiwei classdiscrepancydynamicweightingforcrossdomainfewshothyperspectralimageclassification AT leizhang classdiscrepancydynamicweightingforcrossdomainfewshothyperspectralimageclassification |