Advanced search
1 file | 1.43 MB Add to list

Fully group convolutional neural networks for robust spectral-spatial feature learning

Xian Li (UGent) , Mingli Ding and Aleksandra Pizurica (UGent)
Author
Organization
Abstract
Convolutional neural network (CNN) has been widely applied in hyperspectral image (HSI) classification exhibiting excellent performance. Weak generalization of CNN models to different datasets is a common issue in this domain largely because of limited amount of labeled training samples. In this article, we propose a shape fully group convolutional neural network (FGCNN) method that integrates cascades of shuffled group convolutions tailored to different network stages. To our knowledge, this is the first reported full-group CNN model in general, and we design it in particular for robust spectral-spatial classification of HSI. In the primary feature extraction stage, we develop an original multiscale spectral feature extraction approach based on a novel concept of multi-kernel depthwise convolution that we define in terms of shuffled and importance-weighted group convolution. In the subsequent stage, we introduce a discriminative spectral-spatial feature extraction method with a novel group competition block to capture informative features with relatively few parameters. The final feature fusion stage is defined as a novel lightweight group feature fusion method that sharply reduces fusion weights compared to traditional methods with fully connected layers. Experimental results on three datasets show that the proposed FGCNN yields robust classification accuracy under the same hyperparameter settings compared to the current state-of-the-art.
Keywords
Feature extraction, Hyperspectral imaging, Correlation, Dimensionality reduction, Data models, Data mining, Convolutional neural networks

Downloads

  • FINAL VERSION.PDF
    • full text (Accepted manuscript)
    • |
    • open access
    • |
    • PDF
    • |
    • 1.43 MB

Citation

Please use this url to cite or link to this publication:

MLA
Li, Xian, et al. “Fully Group Convolutional Neural Networks for Robust Spectral-Spatial Feature Learning.” IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2021, pp. 1–14, doi:10.1109/TGRS.2021.3091618.
APA
Li, X., Ding, M., & Pizurica, A. (2021). Fully group convolutional neural networks for robust spectral-spatial feature learning. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 1–14. https://doi.org/10.1109/TGRS.2021.3091618
Chicago author-date
Li, Xian, Mingli Ding, and Aleksandra Pizurica. 2021. “Fully Group Convolutional Neural Networks for Robust Spectral-Spatial Feature Learning.” IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 1–14. https://doi.org/10.1109/TGRS.2021.3091618.
Chicago author-date (all authors)
Li, Xian, Mingli Ding, and Aleksandra Pizurica. 2021. “Fully Group Convolutional Neural Networks for Robust Spectral-Spatial Feature Learning.” IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING: 1–14. doi:10.1109/TGRS.2021.3091618.
Vancouver
1.
Li X, Ding M, Pizurica A. Fully group convolutional neural networks for robust spectral-spatial feature learning. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING. 2021;1–14.
IEEE
[1]
X. Li, M. Ding, and A. Pizurica, “Fully group convolutional neural networks for robust spectral-spatial feature learning,” IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, pp. 1–14, 2021.
@article{8714646,
  abstract     = {{Convolutional neural network (CNN) has been widely applied in hyperspectral image (HSI) classification exhibiting excellent performance. Weak generalization of CNN models to different datasets is a common issue in this domain largely because of limited amount of labeled training samples. In this article, we propose a shape fully group convolutional neural network (FGCNN) method that integrates cascades of shuffled group convolutions tailored to different network stages. To our knowledge, this is the first reported full-group CNN model in general, and we design it in particular for robust spectral-spatial classification of HSI. In the primary feature extraction stage, we develop an original multiscale spectral feature extraction approach based on a novel concept of multi-kernel depthwise convolution that we define in terms of shuffled and importance-weighted group convolution. In the subsequent stage, we introduce a discriminative spectral-spatial feature extraction method with a novel group competition block to capture informative features with relatively few parameters. The final feature fusion stage is defined as a novel lightweight group feature fusion method that sharply reduces fusion weights compared to traditional methods with fully connected layers. Experimental results on three datasets show that the proposed FGCNN yields robust classification accuracy under the same hyperparameter settings compared to the current state-of-the-art.}},
  author       = {{Li, Xian and Ding, Mingli and Pizurica, Aleksandra}},
  issn         = {{0196-2892}},
  journal      = {{IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING}},
  keywords     = {{Feature extraction,Hyperspectral imaging,Correlation,Dimensionality reduction,Data models,Data mining,Convolutional neural networks}},
  language     = {{eng}},
  pages        = {{1--14}},
  title        = {{Fully group convolutional neural networks for robust spectral-spatial feature learning}},
  url          = {{http://dx.doi.org/10.1109/TGRS.2021.3091618}},
  year         = {{2021}},
}

Altmetric
View in Altmetric