Semisupervised Cross-scale Graph Prototypical Network for Hyperspectral Image Classification, TNNLS, 2022.

Bobo Xi, Jiaojiao Li, Yunsong Li, Rui song, Yuchao Xiao, Qian Du and Jocelyn Chanussot.

---

Code for the paper: Semisupervised Cross-scale Graph Prototypical Network for Hyperspectral Image Classification.

Fig. 1: Structure diagram of the proposed X-GPN for HSIC. It comprises four components: multiscale adjacency matrices construction, cross-scale feature learning, SBAA, and a novel prototypical layer.

Training and Test Process

Please run the 'main_IP.py' to reproduce the X-GPN results on IndianPines data set. The training samples distribution and the obtained classification map are shown below. We have successfully test it on Ubuntu 16.04 with Tensorflow 1.13.1 and Keras 2.1.5.

Fig. 2: The composite false-color image, groundtruth, training samples, and classification map of Indian Pines dataset.

Visualization of the feature distribution by t-SNE

Fig. 2: Visualization of the feature distribution obtained by CE, DCE, DCE + TER loss functions on Indian Pines dataset.

References

If you find this code helpful, please kindly cite:

[1] B. Xi, J. Li, Y. Li, R. Song, Y. Xiao, Q. Du, J. Chanussot, “Semisupervised Cross-scale Graph Prototypical Network for Hyperspectral Image Classification,” IEEE Transactions on Neural Networks and Learning Systems, pp. 1-15, 2022, doi:10.1109/TNNLS.2022.3158280.

[2] Y. Li, B. Xi, J. Li, R. Song, Y. Xiao and J. Chanussot, "SGML: A Symmetric Graph Metric Learning Framework for Efficient Hyperspectral Image Classification," in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 15, pp. 609-622, 2022, doi: 10.1109/JSTARS.2021.3135548.

[3] B. Xi, J. Li, Y. Li and Q. Du, "Semi-Supervised Graph Prototypical Networks for Hyperspectral Image Classification," 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS, 2021, pp. 2851-2854, doi: 10.1109/IGARSS47720.2021.9553372.

Citation Details

BibTeX entry:

@ARTICLE{Xi_2022TNNLS_XGPN,
  author={Xi, Bobo and Li, Jiaojiao and Li, Yunsong and Song, Rui and Xiao, Yuchao and Du, Qian and Chanussot, Jocelyn},
  journal={IEEE Transactions on Neural Networks and Learning Systems}, 
  title={Semisupervised Cross-scale Graph Prototypical Network for Hyperspectral Image Classification}, 
  year={2022},
  volume={},
  number={},
  pages={1-15},
  doi={10.1109/TNNLS.2022.3158280}}

@ARTICLE{Xi_2021JSTARS_SGML,
  author={Li, Yunsong and Xi, Bobo and Li, Jiaojiao and Song, Rui and Xiao, Yuchao and Chanussot, Jocelyn},
  journal={IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing}, 
  title={SGML: A Symmetric Graph Metric Learning Framework for Efficient Hyperspectral Image Classification}, 
  year={2022},
  volume={15},
  number={},
  pages={609-622},
  doi={10.1109/JSTARS.2021.3135548}}

@INPROCEEDINGS{Xi_2021IGARSS_GPN,
  author={Xi, Bobo and Li, Jiaojiao and Li, Yunsong and Du, Qian},
  booktitle={2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS}, 
  title={Semi-Supervised Graph Prototypical Networks for Hyperspectral Image Classification}, 
  year={2021},
  volume={},
  number={},
  pages={2851-2854},
  doi={10.1109/IGARSS47720.2021.9553372}}

Licensing

Copyright (C) 2022 Bobo Xi

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 of the License.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program.