{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,28]],"date-time":"2026-02-28T02:39:08Z","timestamp":1772246348257,"version":"3.50.1"},"reference-count":43,"publisher":"MDPI AG","issue":"5","license":[{"start":{"date-parts":[[2020,2,29]],"date-time":"2020-02-29T00:00:00Z","timestamp":1582934400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["61871460"],"award-info":[{"award-number":["61871460"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["61876152"],"award-info":[{"award-number":["61876152"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100012226","name":"Fundamental Research Funds for the Central Universities","doi-asserted-by":"publisher","award":["3102019ghxm016"],"award-info":[{"award-number":["3102019ghxm016"]}],"id":[{"id":"10.13039\/501100012226","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Remote Sensing"],"abstract":"<jats:p>Recently, Hyperspectral Image (HSI) classification methods based on deep learning models have shown encouraging performance. However, the limited numbers of training samples, as well as the mixed pixels due to low spatial resolution, have become major obstacles for HSI classification. To tackle these problems, we propose a resource-efficient HSI classification framework which introduces adaptive spectral unmixing into a 3D\/2D dense network with early-exiting strategy. More specifically, on one hand, our framework uses a cascade of intermediate classifiers throughout the 3D\/2D dense network that is trained end-to-end. The proposed 3D\/2D dense network that integrates 3D convolutions with 2D convolutions is more capable of handling spectral-spatial features, while containing fewer parameters compared with the conventional 3D convolutions, and further boosts the network performance with limited training samples. On another hand, considering the existence of mixed pixels in HSI data, the pixels in HSI classification are divided into hard samples and easy samples. With the early-exiting strategy in these intermediate classifiers, the average accuracy can be improved by reducing the amount of computation cost for easy samples, thus focusing on classifying hard samples. Furthermore, for hard samples, an adaptive spectral unmixing method is proposed as a complementary source of information for classification, which brings considerable benefits to the final performance. Experimental results on four HSI benchmark datasets demonstrate that the proposed method can achieve better performance than state-of-the-art deep learning-based methods and other traditional HSI classification methods.<\/jats:p>","DOI":"10.3390\/rs12050779","type":"journal-article","created":{"date-parts":[[2020,3,3]],"date-time":"2020-03-03T03:13:28Z","timestamp":1583205208000},"page":"779","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":18,"title":["Combining Spectral Unmixing and 3D\/2D Dense Networks with Early-Exiting Strategy for Hyperspectral Image Classification"],"prefix":"10.3390","volume":"12","author":[{"given":"Bei","family":"Fang","sequence":"first","affiliation":[{"name":"School of Computer Science, National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology, Shaanxi Provincial Key Laboratory of Speech Image &amp; Information Processing, Northwestern Polytechnical University, Xi\u2019an 710129, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6923-672X","authenticated-orcid":false,"given":"Yunpeng","family":"Bai","sequence":"additional","affiliation":[{"name":"School of Computing and Information Systems, The University of Melbourne, Melbourne, VIC 3010, Australia"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ying","family":"Li","sequence":"additional","affiliation":[{"name":"School of Computer Science, National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology, Shaanxi Provincial Key Laboratory of Speech Image &amp; Information Processing, Northwestern Polytechnical University, Xi\u2019an 710129, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2020,2,29]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"354","DOI":"10.1109\/JSTARS.2012.2194696","article-title":"Hyperspectral unmixing overview: Geometrical, statistical, and sparse regression-based approaches","volume":"5","author":"Plaza","year":"2012","journal-title":"IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens."},{"key":"ref_2","unstructured":"D\u00f3pido, I.G. (2013). New techniques for hyperspectral image classification. [Ph.D. Thesis, Universidad de Extremadura]."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"652","DOI":"10.1109\/JPROC.2012.2197589","article-title":"Advances in spectral-spatial classification of hyperspectral images","volume":"101","author":"Fauvel","year":"2012","journal-title":"Proc. IEEE"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"1265","DOI":"10.1109\/JSTARS.2019.2903940","article-title":"Abundance-indicated subspace for hyperspectral classification with limited training samples","volume":"12","author":"Xu","year":"2019","journal-title":"IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"44","DOI":"10.1109\/79.974727","article-title":"Spectral unmixing","volume":"19","author":"Keshava","year":"2002","journal-title":"IEEE Signal Proc. Mag."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"S110","DOI":"10.1016\/j.rse.2007.07.028","article-title":"Recent advances in techniques for hyperspectral image processing","volume":"113","author":"Plaza","year":"2009","journal-title":"Remote Sens. Environ."},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Lanaras, C., Baltsavias, E., and Schindler, K. (2015, January 7\u201313). Hyperspectral super-resolution by coupled spectral unmixing. Proceedings of the IEEE International Conference on Computer Vision, Santiago, Chile.","DOI":"10.1109\/ICCV.2015.409"},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"1818","DOI":"10.1109\/TGRS.2015.2489218","article-title":"Coupled sparse denoising and unmixing with low-rank constraint for hyperspectral image","volume":"54","author":"Yang","year":"2015","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"1252","DOI":"10.1109\/LGRS.2015.2390973","article-title":"Informative change detection by unmixing for hyperspectral images","volume":"12","author":"Plaza","year":"2015","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"760","DOI":"10.1109\/LGRS.2011.2109367","article-title":"Unmixing prior to supervised classification of remotely sensed hyperspectral images","volume":"8","author":"Zortea","year":"2011","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"421","DOI":"10.1109\/JSTARS.2011.2176721","article-title":"A quantitative and comparative assessment of unmixing-based feature extraction techniques for hyperspectral image classification","volume":"5","author":"Villa","year":"2012","journal-title":"IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens."},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Alam, F.I., Zhou, J., Tong, L., Liew, A.W.C., and Gao, Y. (December, January 29). Combining unmixing and deep feature learning for hyperspectral image classification. Proceedings of the IEEE 2017 International Conference on Digital Image Computing: Techniques and Applications (DICTA), Sydney, NSW, Australia.","DOI":"10.1109\/DICTA.2017.8227419"},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"4775","DOI":"10.1109\/TGRS.2019.2892903","article-title":"Hyperspectral classification through unmixing abundance maps addressing spectral variability","volume":"57","author":"Drumetz","year":"2019","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"3619","DOI":"10.1109\/JSTARS.2014.2322143","article-title":"A new hybrid strategy combining semisupervised classification and unmixing of hyperspectral data","volume":"7","author":"Li","year":"2014","journal-title":"IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"2899","DOI":"10.1109\/TGRS.2014.2366513","article-title":"Complementarity of discriminative classifiers and spectral unmixing techniques for the interpretation of hyperspectral images","volume":"53","author":"Li","year":"2014","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"1000708","DOI":"10.1117\/12.2241635","article-title":"A new semi-supervised classification strategy combining active learning and spectral unmixing of hyperspectral data","volume":"Volume 10007","author":"Sun","year":"2016","journal-title":"High-Performance Computing in Geoscience and Remote Sensing VI"},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"43","DOI":"10.1016\/j.patcog.2015.08.019","article-title":"Improved hyperspectral image classification by active learning using pre-designed mixed pixels","volume":"51","author":"Samat","year":"2016","journal-title":"Pattern Recognit."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"1975","DOI":"10.1109\/JSTARS.2017.2655516","article-title":"R-VCANet: A new deep-learning-based hyperspectral image classification method","volume":"10","author":"Pan","year":"2017","journal-title":"IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"6232","DOI":"10.1109\/TGRS.2016.2584107","article-title":"Deep feature extraction and classification of hyperspectral images based on convolutional neural networks","volume":"54","author":"Chen","year":"2016","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"847","DOI":"10.1109\/TGRS.2017.2755542","article-title":"Spectral\u2013spatial residual network for hyperspectral image classification: A 3-D deep learning framework","volume":"56","author":"Zhong","year":"2017","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Fang, B., Li, Y., Zhang, H., and Chan, J.C.W. (2019). Hyperspectral images classification based on dense convolutional networks with spectral-wise attention mechanism. Remote Sens., 11.","DOI":"10.3390\/rs11020159"},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"164","DOI":"10.1016\/j.isprsjprs.2020.01.015","article-title":"Collaborative learning of lightweight convolutional neural network and deep clustering for hyperspectral image semi-supervised classification with limited training samples","volume":"161","author":"Fang","year":"2020","journal-title":"ISPRS J. Photogramm. Remote Sens."},{"key":"ref_23","unstructured":"Sutskever, I., Hinton, G.E., and Krizhevsky, A. (2012, January 3\u20138). Imagenet classification with deep convolutional neural networks. Proceedings of the 2012 Advances in Neural Information Processing Systems Conference, Stateline, NV, USA."},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"He, K., Zhang, X., Ren, S., and Sun, J. (2016, January 27\u201330). Deep residual learning for image recognition. Proceedings of the IEEE 2016 Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA.","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"142","DOI":"10.1109\/TPAMI.2015.2437384","article-title":"Region-based convolutional networks for accurate object detection and segmentation","volume":"38","author":"Girshick","year":"2015","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"1915","DOI":"10.1109\/TPAMI.2012.231","article-title":"Learning hierarchical features for scene labeling","volume":"35","author":"Farabet","year":"2012","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"5017","DOI":"10.1109\/TIP.2015.2475625","article-title":"PCANet: A simple deep learning baseline for image classification?","volume":"24","author":"Chan","year":"2015","journal-title":"IEEE Trans. Image Proc."},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"279","DOI":"10.1016\/j.isprsjprs.2019.09.006","article-title":"Deep learning classifiers for hyperspectral imaging: A review","volume":"158","author":"Paoletti","year":"2019","journal-title":"ISPRS J. Photogramm. Remote Sens."},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Wang, W., Dou, S., Jiang, Z., and Sun, L. (2018). A fast dense spectral\u2013spatial convolution network framework for hyperspectral images classification. Remote Sens., 10.","DOI":"10.3390\/rs10071068"},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"1454","DOI":"10.3390\/rs10091454","article-title":"Deep&dense convolutional neural network for hyperspectral image classification","volume":"10","author":"Paoletti","year":"2018","journal-title":"Remote Sens."},{"key":"ref_31","doi-asserted-by":"crossref","unstructured":"Huang, G., Liu, Z., Van Der Maaten, L., and Weinberger, K.Q. (2017, January 21\u201326). Densely connected convolutional networks. Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA.","DOI":"10.1109\/CVPR.2017.243"},{"key":"ref_32","unstructured":"Yu, F., and Koltun, V. (2015). Multi-scale context aggregation by dilated convolutions. arXiv."},{"key":"ref_33","doi-asserted-by":"crossref","unstructured":"K\u00f6p\u00fckl\u00fc, O., Kose, N., Gunduz, A., and Rigoll, G. (2019). Resource efficient 3D convolutional neural networks. arXiv.","DOI":"10.1109\/ICCVW.2019.00240"},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"529","DOI":"10.1109\/36.911111","article-title":"Fully constrained least squares linear spectral mixture analysis method for material quantification in hyperspectral imagery","volume":"39","author":"Heinz","year":"2001","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_35","unstructured":"Huang, G., Chen, D., Li, T., Wu, F., van der Maaten, L., and Weinberger, K.Q. (2017). Multi-scale dense networks for resource efficient image classification. arXiv."},{"key":"ref_36","doi-asserted-by":"crossref","unstructured":"Li, Y., Zhang, H., and Shen, Q. (2017). Spectral\u2013spatial classification of hyperspectral imagery with 3D convolutional neural network. Remote Sens., 9.","DOI":"10.3390\/rs9010067"},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"1949","DOI":"10.1109\/LGRS.2019.2913011","article-title":"Hyperspectral image classification based on 3-D separable ResNet and transfer learning","volume":"16","author":"Jiang","year":"2019","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_38","doi-asserted-by":"crossref","unstructured":"Zhou, Y., Sun, X., Zha, Z.J., and Zeng, W. (2018, January 18\u201322). Mict: Mixed 3d\/2d convolutional tube for human action recognition. Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA.","DOI":"10.1109\/CVPR.2018.00054"},{"key":"ref_39","doi-asserted-by":"crossref","unstructured":"Tran, D., Wang, H., Torresani, L., Ray, J., LeCun, Y., and Paluri, M. (2018, January 18\u201322). A closer look at spatiotemporal convolutions for action recognition. Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA.","DOI":"10.1109\/CVPR.2018.00675"},{"key":"ref_40","doi-asserted-by":"crossref","first-page":"809","DOI":"10.1109\/TGRS.2011.2162649","article-title":"Spectral\u2013spatial hyperspectral image segmentation using subspace multinomial logistic regression and Markov random fields","volume":"50","author":"Li","year":"2011","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_41","doi-asserted-by":"crossref","first-page":"6298","DOI":"10.1109\/TGRS.2013.2296031","article-title":"Spectral\u2013spatial classification of hyperspectral data using local and global probabilities for mixed pixel characterization","volume":"52","author":"Khodadadzadeh","year":"2014","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_42","unstructured":"(2019, December 01). GRSS DASE Website. Available online: http:\/\/dase.grss-ieee.org\/."},{"key":"ref_43","unstructured":"(2019, December 01). 2018 IEEE GRSS Data Fusion Contest. Available online: http:\/\/www.grssieee.org\/community\/technical-committees\/data-fusion."}],"container-title":["Remote Sensing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2072-4292\/12\/5\/779\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T09:02:50Z","timestamp":1760173370000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2072-4292\/12\/5\/779"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2020,2,29]]},"references-count":43,"journal-issue":{"issue":"5","published-online":{"date-parts":[[2020,3]]}},"alternative-id":["rs12050779"],"URL":"https:\/\/doi.org\/10.3390\/rs12050779","relation":{},"ISSN":["2072-4292"],"issn-type":[{"value":"2072-4292","type":"electronic"}],"subject":[],"published":{"date-parts":[[2020,2,29]]}}}