Deep learning provides excellent potentials for hyperspectral images (HSIs) classification, but it is infamous for requiring large amount of labeled samples while the collection of high-quality labels for HSIs is extremely expensive and time-consuming. Furthermore, when the limited training samples are available, deep learning methods may suffer from over-fitting. In this work, we propose a novel collaborative learning framework for semi-supervised HSI classification with joint deep convolutional neural networks and deep clustering. Specifically, a lightweight 3D convolutional neural network (CNN) with much less parameters compared with classical 3D CNNs is designed for deep discriminative feature learning and classification. Then a deep clustering method, that is approximate rank-order clustering (AROC) algorithm, is applied to cluster deep features to generate pseudo labels for abundant unlabeled samples. Finally, we fine-tune the lightweight 3D CNN by minimizing a dual-loss (softmax loss and center loss) using both true and pseudo labels. Experimental results on three challenging HSI datasets demonstrate that the proposed method can achieve better performance than other state-of-the-art deep learning based methods and traditional HSI classification methods methods.
Fang, B, Li, Y, Zhang, H & Chan, JC-W 2020, 'Collaborative learning of lightweight convolutional neural network and deep clustering for hyperspectral image semi-supervised classification with limited training samples', ISPRS Journal of Photogrammetry and Remote Sensing, vol. 161, pp. 164-178. https://doi.org/10.1016/j.isprsjprs.2020.01.015
Fang, B., Li, Y., Zhang, H., & Chan, J. C.-W. (2020). Collaborative learning of lightweight convolutional neural network and deep clustering for hyperspectral image semi-supervised classification with limited training samples. ISPRS Journal of Photogrammetry and Remote Sensing, 161, 164-178. https://doi.org/10.1016/j.isprsjprs.2020.01.015
@article{7f5bfb04e3904b6398c8a02e4ce9d37b,
title = "Collaborative learning of lightweight convolutional neural network and deep clustering for hyperspectral image semi-supervised classification with limited training samples",
abstract = "Deep learning provides excellent potentials for hyperspectral images (HSIs) classification, but it is infamous for requiring large amount of labeled samples while the collection of high-quality labels for HSIs is extremely expensive and time-consuming. Furthermore, when the limited training samples are available, deep learning methods may suffer from over-fitting. In this work, we propose a novel collaborative learning framework for semi-supervised HSI classification with joint deep convolutional neural networks and deep clustering. Specifically, a lightweight 3D convolutional neural network (CNN) with much less parameters compared with classical 3D CNNs is designed for deep discriminative feature learning and classification. Then a deep clustering method, that is approximate rank-order clustering (AROC) algorithm, is applied to cluster deep features to generate pseudo labels for abundant unlabeled samples. Finally, we fine-tune the lightweight 3D CNN by minimizing a dual-loss (softmax loss and center loss) using both true and pseudo labels. Experimental results on three challenging HSI datasets demonstrate that the proposed method can achieve better performance than other state-of-the-art deep learning based methods and traditional HSI classification methods methods.",
author = "Bei Fang and Ying Li and Haokui Zhang and Chan, {Jonathan Cheung-Wai}",
note = "Funding Information: The work was supported in part by the National Natural Science Foundation of China (61871460, 61876152) and Fundamental Research Funds for the Central Universities (3102019ghxm016). The authors would like to thank Prof. P. Gamba from the Pavia University, Pavia, Italy, for providing the reflective optics system imaging spectrometer data and corresponding reference information. The authors would also like to thank the National Center for Airborne Laser Mapping for providing the Houston dataset. The authors would also like to thank Telops Inc. (Qu{\'e}bec, Canada) for acquiring and providing the data used in this study, the IEEE GRSS Image Analysis and Data Fusion Technical Committee and Dr. Michal Shimoni (Signal and Image Centre, Royal Military Academy, Belgium) for organizing the 2014 Data Fusion Contest, the Centre de Recherche Public Gabriel Lippmann (CRPGL, Luxembourg) and Dr. Martin Schlerf (CRPGL) for their contribution of the Hyper-Cam LWIR sensor, and Dr. Michaela De Martino (University of Genoa, Italy) for her contribution to data preparation. The authors would also like to thank Dr. B. Pan from Beihang University for helping with implementing MugNet method. Funding Information: The work was supported in part by the National Natural Science Foundation of China ( 61871460 , 61876152 ) and Fundamental Research Funds for the Central Universities ( 3102019ghxm016 ). Publisher Copyright: {\textcopyright} 2020 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS) Copyright: Copyright 2020 Elsevier B.V., All rights reserved.",
year = "2020",
month = mar,
doi = "10.1016/j.isprsjprs.2020.01.015",
language = "English",
volume = "161",
pages = "164--178",
journal = "ISPRS Journal of Photogrammetry and Remote Sensing",
issn = "0924-2716",
publisher = "Elsevier",
}