EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Lightweight Complex-Valued Siamese Network for Few-Shot PolSAR Image Classification

Yinyin Jiang, Rongzhen Du, Wanying Song, Peng Zhang, Lei Liu, Zhenxi Zhang

Remote Sensing · 20.01.2026 · 18 (2) · 344

DOI: 10.3390/rs18020344 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Remote Sensing, online erschienen am 20.01.2026. Verfasst von Yinyin Jiang und 5 weiteren Autor:innen. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: the Scientific Research Plan Projects of Shannxi Education Department, the Natural Science Basic Research Plan in Shaanxi Province of China, the National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Complex-valued convolutional neural networks (CVCNNs) have demonstrated strong capabilities for polarimetric synthetic aperture radar (PolSAR) image classification by effectively integrating both amplitude and phase information inherent in polarimetric data. However, their practical deployment faces significant challenges due to high computational costs and performance degradation caused by extremely limited labeled samples. To address these challenges, a lightweight CV Siamese network (LCVSNet) is proposed for few-shot PolSAR image classification. Considering the constraints of limited hardware resources in practical applications, simple one-dimensional (1D) CV convolutions along the scattering dimension are combined with two-dimensional (2D) lightweight CV convolutions. In this way, the inter-element dependencies of polarimetric coherency matrix and the spatial correlations between neighboring units can be captured effectively, while simultaneously reducing computational costs. Furthermore, LCVSNet incorporates a contrastive learning (CL) projection head to explicitly optimize the feature space. This optimization can effectively enhance the feature discriminability, leading to accurate classification with a limited number of labeled samples. Experiments on three real PolSAR datasets demonstrate the effectiveness and practical utility of LCVSNet for PolSAR image classification with a small number of labeled samples.

Förderung

  • the Scientific Research Plan Projects of Shannxi Education Department
  • the Natural Science Basic Research Plan in Shaanxi Province of China
  • the National Natural Science Foundation of China
  • the Postdoctoral Fellowship Program of CPSF
  • the Foundation of China for the Central Universities

Zitieren

APA

Jiang, Y., Du, R., Song, W., Zhang, P., Liu, L., & Zhang, Z. (2026). Lightweight Complex-Valued Siamese Network for Few-Shot PolSAR Image Classification. Remote Sensing, 18(2), 344. https://doi.org/10.3390/rs18020344

Vancouver

Jiang Y, Du R, Song W, Zhang P, Liu L, Zhang Z. Lightweight Complex-Valued Siamese Network for Few-Shot PolSAR Image Classification. Remote Sensing. 2026;18(2):344. doi:10.3390/rs18020344

BibTeX
@article{Jiang2026,
  author = {Jiang, Yinyin and Du, Rongzhen and Song, Wanying and Zhang, Peng and Liu, Lei and Zhang, Zhenxi},
  title = {Lightweight Complex-Valued Siamese Network for Few-Shot PolSAR Image Classification},
  journal = {Remote Sensing},
  year = {2026},
  volume = {18},
  number = {2},
  pages = {344},
  doi = {10.3390/rs18020344}
}
Kostenlos starten →