Model‐Inspired Deep Neural Networks for Hyperspectral Unmixing
Yuntao Qian,Fengchao Xiong,Minchao Ye,Jun Zhou
DOI: https://doi.org/10.1002/9781119687788.ch13
2022-01-01
Abstract:Chapter 13 Model-Inspired Deep Neural Networks for Hyperspectral Unmixing Yuntao Qian, Yuntao Qian College of Computer Science, Zhejiang University, Hangzhou, ChinaSearch for more papers by this authorFengchao Xiong, Fengchao Xiong School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing, ChinaSearch for more papers by this authorMinchao Ye, Minchao Ye College of Information Engineering, China Jiliang University, Hangzhou, ChinaSearch for more papers by this authorJun Zhou, Jun Zhou School of Information and Communication Technology, Griffith University, Nathan, QLD, AustraliaSearch for more papers by this author Yuntao Qian, Yuntao Qian College of Computer Science, Zhejiang University, Hangzhou, ChinaSearch for more papers by this authorFengchao Xiong, Fengchao Xiong School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing, ChinaSearch for more papers by this authorMinchao Ye, Minchao Ye College of Information Engineering, China Jiliang University, Hangzhou, ChinaSearch for more papers by this authorJun Zhou, Jun Zhou School of Information and Communication Technology, Griffith University, Nathan, QLD, AustraliaSearch for more papers by this author Book Editor(s):Chein-I Chang, Chein-I Chang Remote Sensing Signal and Image Processing Laboratory, Department of Computer Science and Electrical Engineering, University of Maryland Baltimore County, Baltimore, MD, USASearch for more papers by this author First published: 11 November 2022 https://doi.org/10.1002/9781119687788.ch13 AboutPDFPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShareShare a linkShare onEmailFacebookTwitterLinkedInRedditWechat Summary Model-based and learning-based methods are two typical classes for hyperspectral unmixing. Model-based methods are interpretable but rely on the definition of a physical model and iterative optimization. Learning-based methods have high learning ability, but their network architectures are generic with low physical interpretability. In this chapter, we bridge model-based methods and learning-based methods and introduce model-inspired learning methods. Benefiting from both approaches, model-inspired learning methods are more interpretable with higher learning ability, unmixing efficiency, and effectiveness. References Bioucas-Dias , J.M. , Plaza , A. , Dobigeon , N. et al. ( 2012 ). Hyperspectral unmixing overview: geometrical, statistical, and sparse regression-based approaches . IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens. 5 ( 2 ): 354 – 379 . 10.1109/JSTARS.2012.2194696 Web of Science®Google Scholar Nascimento , J. and Bioucas-Dias , J. ( 2005 ). Vertex component analysis: a fast algorithm to unmix hyperspectral data . IEEE Trans. Geosci. Remote Sens. 43 ( 4 ): 898 – 910 . 10.1109/TGRS.2005.844293 Web of Science®Google Scholar Iordache , M.-D. , Bioucas-Dias , J. , and Plaza , A. (June 2011 ). Sparse unmixing of hyperspectral data . IEEE Trans. Geosci. Remote Sens. 49 ( 6 ): 2014 – 2039 . 10.1109/TGRS.2010.2098413 Web of Science®Google Scholar Heinz , D.C. and Chang , C.-I ( 2001 ). Fully constrained least squares linear spectral mixture analysis method for material quantification in hyperspectral imagery . IEEE Trans. Geosci. Remote Sens. 39 ( 3 ): 529 – 545 . 10.1109/36.911111 Web of Science®Google Scholar Daubechies , I. , Defrise , M. , and De Mol , C. ( 2004 ). An iterative thresholding algorithm for linear inverse problems with a sparsity constraint . Commun. Pure Appl. Math. 57 ( 11 ): 1413 – 1457 . 10.1002/cpa.20042 Web of Science®Google Scholar Beck , A. and Teboulle , M. ( 2009 ). A fast iterative shrinkage-thresholding algorithm for linear inverse problems . SIAM J. Imag. Sci. 2 ( 1 ): 183 – 202 . 10.1137/080716542 Web of Science®Google Scholar Jia , S. and Qian , Y. (Jan. 2009 ). Constrained nonnegative matrix factorization for hyperspectral unmixing . IEEE Trans. Geosci. Remote Sens. 47 ( 1 ): 161 – 173 . 10.1109/TGRS.2008.2002882 Web of Science®Google Scholar Tong , L. , Zhou , J. , Li , X. et al. (Apr. 2017 ). Region-based structure preserving nonnegative matrix factorization for hyperspectral unmixing . IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens. 10 ( 4 ): 1575 – 1588 . 10.1109/JSTARS.2016.2621003 Web of Science®Google Scholar Huang , R. , Li , X. , and Zhao , L. (Oct. 2019 ). Spectral-spatial robust nonnegative matrix factorization for hyperspectral unmixing . IEEE Trans. Geosci. Remote Sens. 57 ( 10 ): 8235 – 8254 . 10.1109/TGRS.2019.2919166 Web of Science®Google Scholar Qian , Y. , Xiong , F. , Zeng , S. et al. ( 2017 ). Matrix-vector nonnegative tensor factorization for blind unmixing of hyperspectral imagery . IEEE Trans. Geosci. Remote Sens. 55 ( 3 ): 1776 – 1792 . 10.1109/TGRS.2016.2633279 Web of Science®Google Scholar Xiong , F. , Qian , Y. , Zhou , J. , and Tang , Y.Y. ( 2019 ). Hyperspectral unmixing via total variation regularized nonnegative tensor factorization . IEEE Trans. Geosci. Remote Sens. 57 ( 4 ): 2341 – 2357 . 10.1109/TGRS.2018.2872888 Web of Science®Google Scholar Boyd , S. , Parikh , N. , Chu , E. et al. ( 2011 ). Distributed optimization and statistical learning via the alternating direction method of multipliers . Found. Trends Mach. Learn. 3 ( 1 ): 1 – 122 . 10.1561/2200000016 PubMedWeb of Science®Google Scholar Wang , W. and Qian , Y. ( 2015 ). Adaptive L 1/2 sparsity-constrained NMF with half-thresholding algorithm for hyperspectral unmixing . IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens. 8 ( 6 ): 2618 – 2631 . 10.1109/JSTARS.2015.2401603 Web of Science®Google Scholar Khajehrayeni , F. and Ghassemian , H. ( 2020 ). Hyperspectral unmixing using deep convolutional autoencoders in a supervised scenario . IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens. 13 : 567 – 576 . 10.1109/JSTARS.2020.2966512 Web of Science®Google Scholar Zhou , L. , Xiao , B. , Liu , X. et al. ( 2019 ). Latent distribution preserving deep subspace clustering . Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI) , Macao, China (10–16 August 2019), 4440 – 4446 . Google Scholar Zhang , X. , Sun , Y. , Zhang , J. et al. ( 2018 ). Hyperspectral unmixing via deep convolutional neural networks . IEEE Geosci. Remote Sens. Lett. 15 ( 11 ): 1755 – 1759 . 10.1109/LGRS.2018.2857804 Web of Science®Google Scholar Su , Y. , Li , J. , Plaza , A. et al. ( Jul. 2019 ). DAEN: deep autoencoder networks for hyperspectral unmixing . IEEE Trans. Geosci. Remote Sens. 57 ( 7 ): 4309 – 4321 . 10.1109/TGRS.2018.2890633 Web of Science®Google Scholar Ozkan , S. , Kaya , B. , and Akar , G.B. ( Jan. 2019 ). EndNet: sparse autoencoder network for endmember extraction and hyperspectral unmixing . IEEE Trans. Geosci. Remote Sens. 57 ( 1 ): 482 – 496 . 10.1109/TGRS.2018.2856929 Web of Science®Google Scholar Min , A. , Guo , Z. , Li , H. , and Peng , J. ( 2022 ). JMnet: joint metric neural network for hyperspectral unmixing . IEEE Trans. Geosci. Remote Sens. 60 : 1 – 12 . 10.1109/TGRS.2022.3160450 Web of Science®Google Scholar Dou , Z. , Gao , K. , Zhang , X. et al. ( 2020 ). Hyperspectral unmixing using orthogonal sparse prior-based autoencoder with hyper-laplacian loss and data-driven outlier detection . IEEE Trans. Geosci. Remote Sens. 58 ( 9 ): 6550 – 6564 . 10.1109/TGRS.2020.2977819 Web of Science®Google Scholar Qu , Y. and Qi , H. ( 2019 ). uDAS: an untied denoising autoencoder with sparsity for spectral unmixing . IEEE Trans. Geosci. Remote Sens. 57 ( 3 ): 1698 – 1712 . 10.1109/TGRS.2018.2868690 Web of Science®Google Scholar Palsson , B. , Ulfarsson , M.O. , and Sveinsson , J.R. ( 2021 ). Convolutional autoencoder for spectralspatial hyperspectral unmixing . IEEE Trans. Geosci. Remote Sens. 59 ( 1 ): 535 – 549 . 10.1109/TGRS.2020.2992743 Web of Science®Google Scholar Hong , D. , Gao , L. , Yao , J. et al. ( 2021 ). Endmember-guided unmixing network (EGU-Net): a general deep learning framework for self-supervised hyperspectral unmixing . IEEE Trans. Neural Netw. Learn. Syst . 1 – 14 . Google Scholar Gao , L. , Han , Z. , Hong , D. et al. ( 2022 ). CyCU-Net: cycle-consistency unmixing network by learning cascaded autoencoders . IEEE Trans. Geosci. Remote Sens. 60 : 1 – 14 . Web of Science®Google Scholar Su , Y. , Xu , X. , Li , J. et al. ( 2021 ). Deep autoencoders with multitask learning for bilinear hyperspectral unmixing . IEEE Trans. Geosci. Remote Sens. 59 ( 10 ): 8615 – 8629 . 10.1109/TGRS.2020.3041157 Web of Science®Google Scholar Zhao , M. , Wang , M. , Chen , J. , and Rahardja , S. ( 2022 ). Hyperspectral unmixing for additive nonlinear models with a 3-D-CNN autoencoder network . IEEE Trans. Geosci. Remote Sens. 60 : 1 – 15 . Web of Science®Google Scholar Zhao , M. , Yan , L. , and Chen , J. ( 2021 ). LSTM-DNN based autoencoder network for nonlinear hyperspectral image unmixing . IEEE J. Sel. Topics Signal Process. 15 ( 2 ): 295 – 309 . 10.1109/JSTSP.2021.3052361 Web of Science®Google Scholar Zhao , M. , Wang , X. , Chen , J. , and Chen , W. ( 2022 ). A plug-and-play priors framework for hyperspectral unmixing . IEEE Trans. Geosci. Remote Sens. 60 : 1 – 13 . Web of Science®Google Scholar Rasti , B. , Koirala , B. , Scheunders , P. , and Ghamisi , P. ( 2022 ). UnDIP: hyperspectral unmixing using deep image prior . IEEE Trans. Geosci. Remote Sens. 60 : 1 – 15 . Google Scholar Hershey , J.R. , Roux , J.L. , and Weninger , F. ( 2014 ). Deep unfolding: model-based inspiration of novel deep architectures , arXiv:1409.2574 (arXiv preprint). Google Scholar Wisdom , S. , Powers , T. , Pitton , J. , and Atlas , L. ( 2017 ). Building recurrent networks by unfolding iterative thresholding for sequential sparse recovery . Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing , 4346 – 4350 , New Orleans, LA, USA (5–8 March 2017). Google Scholar Dong , W. , Zhou , C. , Wu , F. et al. ( 2021 ). Model-guided deep hyperspectral image super-resolution . IEEE Trans. Image Process. 30 : 5754 – 5768 . 10.1109/TIP.2021.3078058 PubMedWeb of Science®Google Scholar He , J. , Li , J. , Yuan , Q. et al. ( 2021 ). Spectral response function-guided deep optimization-driven network for spectral super-resolution . IEEE Trans. Neural Netw. Learn. Syst . 1 – 15 . Google Scholar Shen , D. , Liu , J. , Wu , Z. et al. ( 2022 ). ADMM-HFNet: a matrix decomposition-based deep approach for hyperspectral image fusion . IEEE Trans. Geosci. Remote Sens. 60 : 1 – 17 . Web of Science®Google Scholar Cao , X. , Fu , X. , Hong , D. et al. ( 2022 ). PanCSC-Net: a model-driven deep unfolding method for pansharpening . IEEE Trans. Geosci. Remote Sens. 60 : 1 – 13 . CASWeb of Science®Google Scholar Wang , H. , Xie , Q. , Zhao , Q. , and Meng , D. ( 2020 ). A model-driven deep neural network for single image rain removal . Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (14–19 June 2020). Virtual. Google Scholar Liu , D. , Wang , Z. , Wen , B. et al. ( 2016 ). Robust single image super-resolution via deep networks with sparse prior . IEEE Trans. Image Process. 25 ( 7 ): 3194 – 3207 . 10.1109/TIP.2016.2564643 PubMedWeb of Science®Google Scholar Li , Y. , Tofighi , M. , Geng , J. et al. ( 2020 ). Efficient and interpretable deep blind image deblurring via algorithm unrolling . IEEE Trans. Comput. Imaging 6 : 666 – 681 . 10.1109/TCI.2020.2964202 Web of Science®Google Scholar Gregor , K. and LeCun , Y. ( 2010 ). Learning fast approximations of sparse coding . Proceedings of the International Conference on Machine Learning (ICML) , Haifa, Israel (21–25 June 2010), 399 – 406 . Google Scholar Qian , Y. , Jia , S. , Zhou , J. , and Robles-Kelly , A. ( 2011 ). Hyperspectral unmixing via L 1/2 sparsity-constrained nonnegative matrix factorization . IEEE Trans. Geosci. Remote Sens. 49 ( 11 ): 4282 – 4297 . 10.1109/TGRS.2011.2144605 Web of Science®Google Scholar Xiong , F. , Zhou , J. , Ye , M. et al. ( 2021 ). Nmf-sae: an interpretable sparse autoencoder for hyperspectral unmixing . Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) , Toronto, Canada (6–11 June 2021). Google Scholar He , W. , Zhang , H. , and Zhang , L. ( 2017 ). Total variation regularized reweighted sparse nonnegative matrix factorization for hyperspectral unmixing . IEEE Trans. Geosci. Remote Sens. 55 ( 7 ): 3909 – 3921 . 10.1109/TGRS.2017.2683719 Web of Science®Google Scholar Qian , Y. , Xiong , F. , Qian , Q. , and Zhou , J. ( 2020 ). Spectral mixture model inspired network architectures for hyperspectral unmixing . IEEE Trans. Geosci. Remote Sens. 58 ( 10 ): 7418 – 7434 . 10.1109/TGRS.2020.2982490 Web of Science®Google Scholar Xiong , F. , Zhou , J. , Tao , S. et al. ( 2022 ). Snmf-net: learning a deep alternating neural network for hyperspectral unmixing . IEEE Trans. Geosci. Remote Sens. 60 : 1 – 16 . Web of Science®Google Scholar Lin , C.-J. ( 2007 ). Projected gradient methods for nonnegative matrix factorization . Neural Comput. 19 ( 10 ): 2756 – 2779 . 10.1162/neco.2007.19.10.2756 PubMedWeb of Science®Google Scholar Zuo , W. , Meng , D. , Zhang , L. et al. ( 2013 ). A generalized iterated shrinkage algorithm for non-convex sparse coding . Proceedings of the International Conference on Computer Vision (ICCV) , Sydney, Australia (8–12 April 2013), 217 – 224 . Google Scholar Szegedy , C. , Liu , W. , Jia , Y. et al. ( 2015 ). Going deeper with convolutions . Proceedings of the IEEE Conference on the Computer Vision and Pattern Recognition (CVPR) , Boston, USA (7–12 June 2015), 1 – 9 . Google Scholar Liu , J. and Chen , X. ( 2019 ). ALISTA: analytic weights are as good as learned weights in LISTA . Proceedings of the International Conference on Learning Representations (ICLR) , New Orleans, USA (6–9 May 2019). Google Scholar Salehani , Y.E. and Gazor , S. ( 2017 ). Smooth and sparse regularization for NMF hyperspectral unmixing . IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens. 10 ( 8 ): 3677 – 3692 . 10.1109/JSTARS.2017.2684132 Web of Science®Google Scholar Heylen , R. , Burazerovic , D. , and Scheunders , P. ( 2011 ). Fully constrained least squares spectral unmixing by simplex projection . IEEE Trans. Geosci. Remote Sens. 49 ( 11 ): 4112 – 4122 . 10.1109/TGRS.2011.2155070 Web of Science®Google Scholar Borgerding , M. , Schniter , P. , and Rangan , S. ( 2017 ). AMP-inspired deep networks for sparse linear inverse problems . IEEE Trans. Signal Process. 65 ( 16 ): 4293 – 4308 . 10.1109/TSP.2017.2708040 Web of Science®Google Scholar Advances in Hyperspectral Image Processing Techniques ReferencesRelatedInformation