{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T01:06:35Z","timestamp":1786496795274,"version":"build-2736575974"},"reference-count":38,"publisher":"MDPI AG","issue":"10","license":[{"start":{"date-parts":[[2018,10,11]],"date-time":"2018-10-11T00:00:00Z","timestamp":1539216000000},"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":["61501018"],"award-info":[{"award-number":["61501018"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Remote Sensing"],"abstract":"<jats:p>Deep convolutional neural networks (CNN) have been recently applied to synthetic aperture radar (SAR) for automatic target recognition (ATR) and have achieved state-of-the-art results with significantly improved recognition performance. However, the training period of deep CNN is long, and the size of the network is huge, sometimes reaching hundreds of megabytes. These two factors of deep CNN hinders its practical implementation and deployment in real-time SAR platforms that are typically resource-constrained. To address this challenge, this paper presents three strategies of network compression and acceleration to decrease computing and memory resource dependencies while maintaining a competitive accuracy. First, we introduce a new weight-based network pruning and adaptive architecture squeezing method to reduce the network storage and the time of inference and training process, meanwhile maintain a balance between compression ratio and classification accuracy. Then we employ weight quantization and coding to compress the network storage space. Due to the fact that the amount of calculation is mainly reflected in the convolution layer, a fast approach for pruned convolutional layers is proposed to reduce the number of multiplication by exploiting the sparsity in the activation inputs and weights. Experimental results show that the convolutional neural networks for SAR-ATR can be compressed by     40 \u00d7     without loss of accuracy, and the number of multiplication can be reduced by     15 \u00d7    . Combining these strategies, we can easily load the network in resource-constrained platforms, speed up the inference process to get the results in real-time or even retrain a more suitable network with new image data in a specific situation.<\/jats:p>","DOI":"10.3390\/rs10101618","type":"journal-article","created":{"date-parts":[[2018,10,12]],"date-time":"2018-10-12T02:58:04Z","timestamp":1539313084000},"page":"1618","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":47,"title":["Slim and Efficient Neural Network Design for Resource-Constrained SAR Target Recognition"],"prefix":"10.3390","volume":"10","author":[{"given":"Hongyi","family":"Chen","sequence":"first","affiliation":[{"name":"College of Information Science and Technology, Beijing University of Chemical Technology, Beijing 100029, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2058-2373","authenticated-orcid":false,"given":"Fan","family":"Zhang","sequence":"additional","affiliation":[{"name":"College of Information Science and Technology, Beijing University of Chemical Technology, Beijing 100029, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bo","family":"Tang","sequence":"additional","affiliation":[{"name":"Department of Electrical and Computer Engineering, Mississippi State University, Starkville, MS 39759, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Qiang","family":"Yin","sequence":"additional","affiliation":[{"name":"College of Information Science and Technology, Beijing University of Chemical Technology, Beijing 100029, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xian","family":"Sun","sequence":"additional","affiliation":[{"name":"Institute of Electronics, Chinese Academy of Sciences, Beijing 100190, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2018,10,11]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"1225","DOI":"10.1109\/TGRS.2014.2336241","article-title":"A High-Order Imaging Algorithm for High-Resolution Spaceborne SAR Based on a Modified Equivalent Squint Range Model","volume":"53","author":"Wang","year":"2014","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"2115","DOI":"10.1109\/JSTARS.2017.2787728","article-title":"Multiple Mode SAR Raw Data Simulation and Parallel Acceleration for Gaofen-3 Mission","volume":"11","author":"Zhang","year":"2018","journal-title":"IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens."},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"Matuszewski, J., and Sikorska-\u0141ukasiewicz, K. (2017, January 28\u201330). Neural network application for emitter identification. Proceedings of the 2017 18th International Radar Symposium (IRS), Prague, Czech Republic.","DOI":"10.23919\/IRS.2017.8008202"},{"key":"ref_4","first-page":"46","article-title":"Objects Detection and Recognition System Using Artificial Neural Networks and Drones","volume":"6","author":"Pietrow","year":"2018","journal-title":"J. Electr. Eng."},{"key":"ref_5","doi-asserted-by":"crossref","unstructured":"Matuszewski, J. (2018, January 20\u201324). Radar signal identification using a neural network and pattern recognition methods. Proceedings of the 2018 14th International Conference on Advanced Trends in Radioelecrtronics, Telecommunications and Computer Engineering (TCSET), Lviv-Slavske, Ukraine.","DOI":"10.1109\/TCSET.2018.8336160"},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"481","DOI":"10.1117\/12.321851","article-title":"Moving and stationary target acquisition and recognition (MSTAR) public dataset. Technical report","volume":"3370","author":"Diemunsch","year":"1995","journal-title":"Proc. SPIE"},{"key":"ref_7","first-page":"3","article-title":"An overview of automatic target recognition","volume":"6","author":"Dudgeon","year":"1993","journal-title":"Lincoln Lab. J."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"220","DOI":"10.1016\/j.neucom.2017.06.004","article-title":"A novel target detection method for SAR images based on shadow proposal and saliency analysis","volume":"267","author":"Gao","year":"2017","journal-title":"Neurocomputing"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"1092","DOI":"10.1109\/TAES.2013.120378","article-title":"Modified polar mapping classifier for SAR automatic target recognition","volume":"50","author":"Park","year":"2014","journal-title":"IEEE Trans. Aerosp. Electron. Syst."},{"key":"ref_10","first-page":"658","article-title":"A Method of SAR Target Recognition Based on Gabor Filter and Local Texture Feature Extraction","volume":"4","author":"Wang","year":"2015","journal-title":"J. Radars"},{"key":"ref_11","unstructured":"Gao, F., Mei, J., Sun, J., Wang, J., Yang, E., and Hussain, A. (2015). Target detection and recognition in SAR imagery based on KFDA. J. Syst. Eng. Electron., in press."},{"key":"ref_12","unstructured":"Knapskog, A.O. (2010, January 7\u201310). Classification of ships in TerraSAR-X images based on 3D models and silhouette matching. Proceedings of the European Conference on Synthetic Aperture Radar (EUSAR), Aachen, Germany."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"643","DOI":"10.1109\/7.937475","article-title":"Support vector machines for SAR automatic target recognition","volume":"37","author":"Zhao","year":"2001","journal-title":"IEEE Trans. Aerosp. Electron. Syst."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"112","DOI":"10.1109\/TAES.2007.357120","article-title":"Adaptive boosting for SAR automatic target recognition","volume":"43","author":"Sun","year":"2007","journal-title":"IEEE Trans. Aerosp. Electron. Syst."},{"key":"ref_15","unstructured":"Hummel, R. (2000, January 12). Model-based ATR using synthetic aperture radar. Proceedings of the European Conference on Synthetic Aperture Radar (EUSAR), Alexandria, VA, USA."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"1153","DOI":"10.1016\/0893-6080(95)00050-X","article-title":"Neural networks for automatic target recognition","volume":"8","author":"Rogers","year":"1995","journal-title":"Neural Netw."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"4806","DOI":"10.1109\/TGRS.2016.2551720","article-title":"Target Classification Using the Deep Convolutional Networks for SAR Images","volume":"54","author":"Chen","year":"2016","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"2351","DOI":"10.1109\/LGRS.2015.2478256","article-title":"High-Resolution SAR Image Classification via Deep Convolutional Autoencoders","volume":"12","author":"Geng","year":"2015","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_19","unstructured":"Lv, Q., Dou, Y., Niu, X., Xu, J., and Li, B. (2014, January 13\u201318). Classification of land cover based on deep belief networks using polarimetric RADARSAT-2 data. Proceedings of the 2014 IEEE Geoscience and Remote Sensing Symposium, Quebec City, QC, Canada."},{"key":"ref_20","unstructured":"Liu, C., Yin, J., and Yang, J. (2015, January 14\u201316). Application of deep learning to polarimetric SAR classification. Proceedings of the IET International Radar Conference 2015, Hangzhou, China."},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Bentes, C., Velotto, D., and Lehner, S. (2015, January 26\u201331). Target Classification in Oceanographic SAR Images With Deep Neural Networks: Architecture and Initial Results. Proceedings of the IEEE Geoscience and Remote Sensing Symposium (IGARSS), Milan, Italy.","DOI":"10.1109\/IGARSS.2015.7326627"},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Li, X., Li, C., Wang, P., Men, Z., and Xu, H. (2015, January 1\u20134). SAR ATR Based on Dividing CNN into CAE and SNN. Proceedings of the IEEE Asia-Pacific Conference on Synthetic Aperture Radar (APSAR), Singapore.","DOI":"10.1109\/APSAR.2015.7306296"},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"26880","DOI":"10.1109\/ACCESS.2017.2773363","article-title":"Multi-Aspect-Aware Bidirectional LSTM Networks for Synthetic Aperture Radar Target Recognition","volume":"5","author":"Zhang","year":"2017","journal-title":"IEEE Access"},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"2196","DOI":"10.1109\/TGRS.2017.2776357","article-title":"SAR Automatic Target Recognition Based on Multiview Deep Learning Framework","volume":"56","author":"Pei","year":"2018","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"2245","DOI":"10.1109\/LGRS.2017.2758900","article-title":"Zero-shot learning of SAR target feature space with deep generative neural networks","volume":"14","author":"Song","year":"2017","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Li, R., Liu, W., Yang, L., Sun, S., Hu, W., Zhang, F., and Li, W. (2018). DeepUNet: A Deep Fully Convolutional Network for Pixel-Level Sea-Land Segmentation. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 1\u20139.","DOI":"10.1109\/JSTARS.2018.2833382"},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"258619","DOI":"10.1155\/2015\/258619","article-title":"Deep convolutional neural networks for hyperspectral image classification","volume":"2015","author":"Hu","year":"2015","journal-title":"J. Sens."},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"1111","DOI":"10.1109\/LGRS.2017.2699196","article-title":"Synthetic aperture radar image synthesis by using generative adversarial nets","volume":"14","author":"Guo","year":"2017","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Gao, F., Yang, Y., Wang, J., Sun, J., Yang, E., and Zhou, H. (2018). A deep convolutional generative adversarial networks (DCGANs)-based semi-supervised method for object recognition in synthetic aperture radar (SAR) images. Remote Sens., 10.","DOI":"10.3390\/rs10060846"},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Huang, Z., Pan, Z., and Lei, B. (2017). Transfer learning with deep convolutional neural network for SAR target classification with limited labeled data. Remote Sens., 9.","DOI":"10.3390\/rs9090907"},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"1091","DOI":"10.1109\/LGRS.2017.2698213","article-title":"Deep convolutional highway unit network for sar target classification with limited labeled training data","volume":"14","author":"Lin","year":"2017","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_32","unstructured":"Iandola, F.N., Han, S., Moskewicz, M.W., Ashraf, K., Dally, W.J., and Keutzer, K. (arXiv, 2016). SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5 MB model size, arXiv."},{"key":"ref_33","unstructured":"Han, S., Mao, H., and Dally, W.J. (arXiv, 2015). Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding, arXiv."},{"key":"ref_34","unstructured":"Chen, W., Wilson, J.T., Tyree, S., Weinberger, K.Q., and Chen, Y. (arXiv, 2015). Compressing Neural Networks with the Hashing Trick, arXiv."},{"key":"ref_35","unstructured":"Gong, Y., Liu, L., Yang, M., and Bourdev, L. (arXiv, 2014). Compressing Deep Convolutional Networks using Vector Quantization, arXiv."},{"key":"ref_36","unstructured":"Liu, X., Pool, J., Han, S., and Dally, W.J. (arXiv, 2018). Efficient Sparse-Winograd Convolutional Neural Networks, arXiv."},{"key":"ref_37","unstructured":"Han, S., Pool, J., Tran, J., and Dally, W.J. (arXiv, 2015). Learning both Weights and Connections for Efficient Neural Networks, arXiv."},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"554","DOI":"10.1117\/12.321858","article-title":"An evaluation of SAR ATR algorithm performance sensitivity to MSTAR extended operating conditions","volume":"3370","author":"Mossing","year":"1998","journal-title":"Proc. SPIE"}],"container-title":["Remote Sensing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2072-4292\/10\/10\/1618\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T15:25:06Z","timestamp":1760196306000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2072-4292\/10\/10\/1618"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2018,10,11]]},"references-count":38,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2018,10]]}},"alternative-id":["rs10101618"],"URL":"https:\/\/doi.org\/10.3390\/rs10101618","relation":{},"ISSN":["2072-4292"],"issn-type":[{"value":"2072-4292","type":"electronic"}],"subject":[],"published":{"date-parts":[[2018,10,11]]}}}