{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,11,5]],"date-time":"2025-11-05T06:31:40Z","timestamp":1762324300960,"version":"build-2065373602"},"reference-count":47,"publisher":"MDPI AG","issue":"7","license":[{"start":{"date-parts":[[2018,7,21]],"date-time":"2018-07-21T00:00:00Z","timestamp":1532131200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Remote Sensing"],"abstract":"<jats:p>Spectral-spatial classification of hyperspectral images has been the subject of many studies in recent years. When there are only a few labeled pixels for training and a skewed class label distribution, this task becomes very challenging because of the increased risk of overfitting when training a classifier. In this paper, we show that in this setting, a convolutional neural network with a single hidden layer can achieve state-of-the-art performance when three tricks are used: a spectral-locality-aware regularization term and smoothing- and label-based data augmentation. The shallow network architecture prevents overfitting in the presence of many features and few training samples. The locality-aware regularization forces neighboring wavelengths to have similar contributions to the features generated during training. The new data augmentation procedure favors the selection of pixels in smaller classes, which is beneficial for skewed class label distributions. The accuracy of the proposed method is assessed on five publicly available hyperspectral images, where it achieves state-of-the-art results. As other spectral-spatial classification methods, we use the entire image (labeled and unlabeled pixels) to infer the class of its unlabeled pixels. To investigate the positive bias induced by the use of the entire image, we propose a new learning setting where unlabeled pixels are not used for building the classifier. Results show the beneficial effect of the proposed tricks also in this setting and substantiate the advantages of using labeled and unlabeled pixels from the image for hyperspectral image classification.<\/jats:p>","DOI":"10.3390\/rs10071156","type":"journal-article","created":{"date-parts":[[2018,7,24]],"date-time":"2018-07-24T02:58:56Z","timestamp":1532401136000},"page":"1156","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":34,"title":["Spectral-Spatial Classification of Hyperspectral Images: Three Tricks and a New Learning Setting"],"prefix":"10.3390","volume":"10","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-5870-4973","authenticated-orcid":false,"given":"Jacopo","family":"Acquarelli","sequence":"first","affiliation":[{"name":"Institute for Computing and Information Science, Radboud University Nijmegen, 6525 EC Nijmegen, The Netherlands"},{"name":"Institute for Molecules and Materials, Radboud University Nijmegen, 6525 AJ Nijmegen, The Netherlands"}]},{"given":"Elena","family":"Marchiori","sequence":"additional","affiliation":[{"name":"Institute for Computing and Information Science, Radboud University Nijmegen, 6525 EC Nijmegen, The Netherlands"}]},{"given":"Lutgarde","family":"Buydens","sequence":"additional","affiliation":[{"name":"Institute for Molecules and Materials, Radboud University Nijmegen, 6525 AJ Nijmegen, The Netherlands"}]},{"given":"Thanh","family":"Tran","sequence":"additional","affiliation":[{"name":"Institute for Computing and Information Science, Radboud University Nijmegen, 6525 EC Nijmegen, The Netherlands"},{"name":"Institute for Molecules and Materials, Radboud University Nijmegen, 6525 AJ Nijmegen, The Netherlands"},{"name":"Corbion, 4206 AC Gorinchem, The Netherlands"}]},{"given":"Twan","family":"Van Laarhoven","sequence":"additional","affiliation":[{"name":"Institute for Computing and Information Science, Radboud University Nijmegen, 6525 EC Nijmegen, The Netherlands"},{"name":"Faculty of Management, Science and Technology, Open University of The Netherlands, 6419 AT Heerlen, The Netherlands"}]}],"member":"1968","published-online":{"date-parts":[[2018,7,21]]},"reference":[{"key":"ref_1","unstructured":"Chang, C.I. (2003). Hyperspectral Imaging: Techniques for Spectral Detection and Classification, Springer."},{"key":"ref_2","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":"2013","journal-title":"Proc. IEEE"},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"45","DOI":"10.1109\/MSP.2013.2279179","article-title":"Advances in hyperspectral image classification: Earth monitoring with statistical learning methods","volume":"31","author":"Tuia","year":"2014","journal-title":"IEEE Sign. Proc. Mag."},{"key":"ref_4","first-page":"402","article-title":"Unsupervised Full-Polarimetric SAR Data Segmentation as a Tool for Classification of Agricultural Areas","volume":"4","author":"Hoekman","year":"2011","journal-title":"IEEE J. STARS"},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1016\/j.chemolab.2004.07.011","article-title":"Clustering multispectral images: A tutorial","volume":"77","author":"Tran","year":"2005","journal-title":"Chemom. Intell. Lab. Syst."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"1912","DOI":"10.1109\/TGRS.2005.848427","article-title":"Initialization of Markov random field clustering of large remote sensing images","volume":"43","author":"Tran","year":"2005","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"5771","DOI":"10.1109\/TGRS.2013.2292544","article-title":"Automatic spectral\u2014Spatial classification framework based on attribute profiles and supervised feature extraction","volume":"52","author":"Ghamisi","year":"2014","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"6223","DOI":"10.1109\/TGRS.2015.2436335","article-title":"Spectral and spatial classification of hyperspectral images based on ICA and reduced morphological attribute profiles","volume":"53","author":"Falco","year":"2015","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"2696","DOI":"10.1109\/TGRS.2014.2363682","article-title":"Spectral\u2014Spatial classification of hyperspectral images via spatial translation-invariant wavelet-based sparse representation","volume":"53","author":"He","year":"2015","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_10","doi-asserted-by":"crossref","unstructured":"Hu, F., Xia, G.S., Wang, Z., Huang, X., Zhang, L., and Sun, H. (2015). Unsupervised feature learning via spectral clustering of multidimensional patches for remotely sensed scene classification. IEEE J. STARS, 8.","DOI":"10.1109\/JSTARS.2015.2444405"},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"4472","DOI":"10.1109\/TGRS.2015.2400449","article-title":"Learning high-level features for satellite image classification with limited labeled samples","volume":"53","author":"Yang","year":"2015","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"1381","DOI":"10.1109\/TGRS.2016.2623742","article-title":"Discriminative Low-Rank Gabor Filtering for Spectral\u2014Spatial Hyperspectral Image Classification","volume":"55","author":"He","year":"2017","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"3574","DOI":"10.1109\/TIP.2014.2329767","article-title":"Hyperspectral image segmentation using a new spectral unmixing-based binary partition tree representation","volume":"23","author":"Veganzones","year":"2014","journal-title":"IEEE Trans. Image Process."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"4398","DOI":"10.1109\/TGRS.2017.2691906","article-title":"From Subpixel to Superpixel: A Novel Fusion Framework for Hyperspectral Image Classification","volume":"55","author":"Lu","year":"2017","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"212","DOI":"10.1109\/TGRS.2016.2604290","article-title":"Random-walker-based collaborative learning for hyperspectral image classification","volume":"55","author":"Sun","year":"2017","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"3071","DOI":"10.1109\/TGRS.2017.2650938","article-title":"A Novel Semisupervised Active-Learning Algorithm for Hyperspectral Image Classification","volume":"55","author":"Wang","year":"2017","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"2693","DOI":"10.1109\/TGRS.2017.2651639","article-title":"Self-taught feature learning for hyperspectral image classification","volume":"55","author":"Kemker","year":"2017","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"Lee, H., and Kwon, H. (arXiv, 2016). Contextual Deep CNN Based Hyperspectral Classification, arXiv.","DOI":"10.1109\/IGARSS.2016.7729859"},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"Hu, W., Huang, Y., Wei, L., Zhang, F., and Li, H. (2015). Deep convolutional neural networks for hyperspectral image classification. J. Sens., 2015.","DOI":"10.1155\/2015\/258619"},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Makantasis, K., Karantzalos, K., Doulamis, A., and Doulamis, N. (2015, January 26\u201331). Deep Supervised Learning for Hyperspectral Data Classification through Convolutional Neural Networks. Proceedings of the 2015 IEEE International Geoscience and Remote Sensing Symposium, Milan, Italy.","DOI":"10.1109\/IGARSS.2015.7326945"},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Slavkovikj, V., Verstockt, S., Neve, W.D., Hoecke, S.V., and Walle, R.V.D. (2015, January 26\u201330). Hyperspectral Image Classification with Convolutional Neural Networks. Proceedings of the 23rd ACM international conference on Multimedia, Brisbane, Australia.","DOI":"10.1145\/2733373.2806306"},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Liang, H., and Li, Q. (2016). Hyperspectral imagery classification using sparse representations of convolutional neural network features. Remote Sens., 8.","DOI":"10.3390\/rs8020099"},{"key":"ref_23","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_24","doi-asserted-by":"crossref","first-page":"14680","DOI":"10.3390\/rs71114680","article-title":"Transferring Deep Convolutional Neural Networks for the Scene Classification of High-Resolution Remote Sensing Imagery","volume":"7","author":"Hu","year":"2015","journal-title":"Remote Sens."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"127","DOI":"10.1080\/10095020.2017.1399674","article-title":"Object-based classification of hyperspectral data using Random Forest algorithm","volume":"21","author":"Amini","year":"2018","journal-title":"Geo-Spat. Inf. Sci."},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"O\u2019Neil-Dunne, J., Pelletier, K., MacFaden, S., Troy, A., and Grove, J.M. (2009, January 12\u201314). Object-based high-resolution land-cover mapping. Proceedings of the 2009 17th International Conference on Geoinformatics, Fairfax, VA, USA.","DOI":"10.1109\/GEOINFORMATICS.2009.5293435"},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Blaschke, T., Lang, S., and Hay, G.J. (2008). Object-Based Image Analysis: Spatial Concepts for Knowledge-Driven Remote Sensing Applications, Springer. Lecture Notes in Geoinformation and Cartography.","DOI":"10.1007\/978-3-540-77058-9"},{"key":"ref_28","first-page":"434","article-title":"Classification of Hyperspectral Remote Sensing Image Using Hierarchical Local-Receptive-Field-Based Extreme Learning Machine","volume":"13","author":"Lv","year":"2016","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"6547","DOI":"10.1109\/TGRS.2017.2729882","article-title":"Multiple Kernel Learning for Hyperspectral Image Classification: A Review","volume":"55","author":"Gu","year":"2017","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"2117","DOI":"10.1109\/LGRS.2017.2753401","article-title":"Hyperspectral Image Classification via Low-Rank and Sparse Representation With Spectral Consistency Constraint","volume":"14","author":"Pan","year":"2017","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"102","DOI":"10.1016\/j.patcog.2016.09.011","article-title":"Probabilistic class structure regularized sparse representation graph for semi-supervised hyperspectral image classification","volume":"63","author":"Shao","year":"2017","journal-title":"Pattern Recognit."},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"1579","DOI":"10.1109\/TGRS.2017.2765364","article-title":"Recent Advances on Spectral-Spatial Hyperspectral Image Classification: An Overview and New Guidelines","volume":"56","author":"He","year":"2017","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_33","first-page":"2094","article-title":"Deep Learning-Based Classification of Hyperspectral Data","volume":"7","author":"Chen","year":"2014","journal-title":"IEEE J. STARS"},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"217","DOI":"10.1109\/TGRS.2012.2201730","article-title":"Hyperspectral image classification via kernel sparse representation","volume":"51","author":"Chen","year":"2013","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"5861","DOI":"10.1109\/TGRS.2015.2423688","article-title":"Superpixel-based graphical model for remote sensing image mapping","volume":"53","author":"Zhang","year":"2015","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_36","first-page":"4073","article-title":"Spectral-Spatial Classification of Hyperspectral Image Based on Deep Auto-Encoder","volume":"9","author":"Ma","year":"2016","journal-title":"IEEE J. STARS"},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"88","DOI":"10.1016\/j.neucom.2016.09.010","article-title":"Convolutional neural networks for hyperspectral image classification","volume":"219","author":"Yu","year":"2017","journal-title":"Neurocomputing"},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"4843","DOI":"10.1109\/TIP.2017.2725580","article-title":"Going Deeper With Contextual CNN for Hyperspectral Image Classification","volume":"26","author":"Lee","year":"2017","journal-title":"IEEE Trans. Image Process."},{"key":"ref_39","doi-asserted-by":"crossref","first-page":"862","DOI":"10.1109\/TGRS.2016.2616489","article-title":"On the sampling strategy for evaluation of spectral-spatial methods in hyperspectral image classification","volume":"55","author":"Liang","year":"2017","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_40","unstructured":"Glorot, X., and Bengio, Y. (2010, January 13\u201315). Understanding the difficulty of training deep feedforward neural networks. Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics (AISTATS 2010), Sardinia, Italy."},{"key":"ref_41","unstructured":"Lechevallier, Y., and Saporta, G. (2010, January 22\u201327). Large-Scale Machine Learning with Stochastic Gradient Descent. Proceedings of the 19th International Conference on Computational Statistics, COMPSTAT\u20192010, Paris, France."},{"key":"ref_42","doi-asserted-by":"crossref","first-page":"22","DOI":"10.1016\/j.aca.2016.12.010","article-title":"Convolutional neural networks for vibrational spectroscopic data analysis","volume":"954","author":"Acquarelli","year":"2017","journal-title":"Anal. Chim. Acta"},{"key":"ref_43","doi-asserted-by":"crossref","first-page":"297","DOI":"10.1109\/LGRS.2009.2012443","article-title":"Improving Hyperspectral Image Classification Using Spatial Preprocessing","volume":"6","author":"Manian","year":"2009","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_44","first-page":"281","article-title":"Random Search for Hyper-parameter Optimization","volume":"13","author":"Bergstra","year":"2012","journal-title":"J. Machine Learn. Res."},{"key":"ref_45","doi-asserted-by":"crossref","first-page":"317","DOI":"10.1023\/A:1009752403260","article-title":"On Comparing Classifiers: Pitfalls to Avoid and a Recommended Approach","volume":"1","author":"Salzberg","year":"1997","journal-title":"Data Min. Knowl. Discov."},{"key":"ref_46","doi-asserted-by":"crossref","first-page":"3742","DOI":"10.1109\/TGRS.2013.2275613","article-title":"Feature Extraction of Hyperspectral Images With Image Fusion and Recursive Filtering","volume":"52","author":"Kang","year":"2014","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_47","unstructured":"Cope, B. (2006). Implementation of 2D Convolution on FPGA, GPU and CPU, Imperial College Report."}],"container-title":["Remote Sensing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2072-4292\/10\/7\/1156\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T15:13:33Z","timestamp":1760195613000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2072-4292\/10\/7\/1156"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2018,7,21]]},"references-count":47,"journal-issue":{"issue":"7","published-online":{"date-parts":[[2018,7]]}},"alternative-id":["rs10071156"],"URL":"https:\/\/doi.org\/10.3390\/rs10071156","relation":{},"ISSN":["2072-4292"],"issn-type":[{"type":"electronic","value":"2072-4292"}],"subject":[],"published":{"date-parts":[[2018,7,21]]}}}