EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

ECSA-UNet: End-to-End Compressive Spatial-Spectral Attention UNet for Hyperspectral Image Classification

Hao Zhang, Yusen Liu, Chao Wen, Xianhong Zhao, Fei Han, Tingting He, Enlong Wang, Yicheng Wang, Shouzheng Zhu, Genghua Huang, Fashuai Li, Yuwei Chen

Remote Sensing · 04.11.2026 · 18 (21) · 3790

DOI: 10.3390/rs18213790 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Remote Sensing, online erschienen am 04.11.2026. Verfasst von Hao Zhang und 11 weiteren Autor:innen. Die Arbeit nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: the Youth Independent Innovation Science Fund.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Hyperspectral image (HSI) classification is a fundamental remote sensing task, but the high-dimensional nature of HSI data poses challenges for acquisition, storage, transmission, and processing. To mitigate these issues, researchers have developed compressive classification methods based on the coded aperture snapshot spectral imaging (CASSI) system, enabling direct classification in the compressed domain without reconstructing the full hyperspectral cube. However, existing methods often fail to fully leverage spatial-spectral correlations, suffer from limited multi-scale feature extraction, and are constrained by fixed receptive fields and degradation caused by compression. To overcome these limitations, we propose ECSA-UNet, an end-to-end framework that directly classifies compressive spectral measurements from CASSI. The model introduces a spatial-spectral decoupling module and refines the U-Net architecture with global–local feature extraction, Spatial-Spectral Attention, a Multi-Scale Enhancement Pyramid and Cross-Stage Connections, enabling dynamic receptive fields, improved long-range dependency modeling, and robust local–global feature fusion. Moreover, the CASSI coded aperture is parameterized as a trainable component, allowing for joint optimization of coding and classification within an end-to-end learning paradigm. Extensive experiments on four benchmarks show ECSA-UNet achieves state-of-the-art accuracy, and ablations confirm each module’s contribution. In addition, it maintains an effective trade-off between parameter overhead and computational efficiency, offering a viable solution for latency-sensitive hyperspectral remote sensing tasks.

Förderung

  • the Youth Independent Innovation Science Fund

Zitieren

APA

Zhang, H., Liu, Y., Wen, C., Zhao, X., Han, F., He, T., Wang, E., Wang, Y., Zhu, S., Huang, G., Li, F., & Chen, Y. (2026). ECSA-UNet: End-to-End Compressive Spatial-Spectral Attention UNet for Hyperspectral Image Classification. Remote Sensing, 18(21), 3790. https://doi.org/10.3390/rs18213790

Vancouver

Zhang H, Liu Y, Wen C, Zhao X, Han F, He T, et al. ECSA-UNet: End-to-End Compressive Spatial-Spectral Attention UNet for Hyperspectral Image Classification. Remote Sensing. 2026;18(21):3790. doi:10.3390/rs18213790

BibTeX
@article{Zhang2026,
  author = {Zhang, Hao and Liu, Yusen and Wen, Chao and Zhao, Xianhong and Han, Fei and He, Tingting and Wang, Enlong and Wang, Yicheng and Zhu, Shouzheng and Huang, Genghua and Li, Fashuai and Chen, Yuwei},
  title = {ECSA-UNet: End-to-End Compressive Spatial-Spectral Attention UNet for Hyperspectral Image Classification},
  journal = {Remote Sensing},
  year = {2026},
  volume = {18},
  number = {21},
  pages = {3790},
  doi = {10.3390/rs18213790}
}
Kostenlos starten →