Deep Learning in Maritime Autonomous Surface Ships: Current Development and Challenges

Abstract

Autonomous surface ships have become increasingly interesting for commercial maritime sectors. Before deep learning (DL) was proposed, surface ship autonomy was mostly model-based. The development of artificial intelligence (AI) has prompted new challenges in the maritime industry. A detailed literature study and examination of DL applications in autonomous surface ships are still missing. Thus, this article reviews the current progress and applications of DL in the field of ship autonomy. The history of different DL methods and their application in autonomous surface ships is briefly outlined. Then, the previously published works studying DL methods in ship autonomy have been categorized into four groups, i.e., control systems, ship navigation, monitoring system, and transportation and logistics. The state-of-the-art of this review paper majorly lies in presenting the existing limitations and innovations of different applications. Subsequently, the current issues and challenges for DL application in autonomous surface ships are discussed. In addition, we have proposed a comparative study of traditional and DL algorithms in ship autonomy and also provided the future research scope as well.

Keywords

Maritime autonomous surface ships / Deep learning (DL) / Artificial intelligence (AI) / Review

References

[1] Albarqouni S, Baur C, Achilles F, Belagiannis V, Navab N. Aggnet: Deep learning from crowds for mitosis detection in breast cancer histology images. IEEE Transactions on Medical Imaging, 2016, 35(5): 1313-1321

[2] Ammour N, Alhichri H, Bazi Y, Benjdira B, Alajlan N, Zuair M. Deep learning approach for car detection in UAV imagery. Remote Sensing, 2017, 9(4): 312

[3] Bentes C, Velotto D, Tings B. Ship classification in terrasar-x images with convolutional neural networks. IEEE Journal of Oceanic Engineering, 2018, 43(1): 258-266

[4] Bentes C, Frost A, Velotto D, Tings B (2016) Ship-iceberg discrimination with convolutional neural networks in high resolution SAR images. In Proceedings of EUSAR 2016: 11th European Conference on Synthetic Aperture Radar, 1–4

[5] Cao X, Shu G, Chen L, Wang Y (2020) Ship recognition method combined with image segmentation and deep learning feature extraction in video surveillance. Multimedia Tools and Applications, 79. https://doi.org/10.1007/s11042-018-7138-3

[6] Chang YL, Anagaw A, Chang L, Wang Y, Hsiao CY, Lee WH. Ship detection based on yolov2 for sar imagery. Remote Sensing, 2019, 11(7): 786

[7] Chen C, Chen XQ, Ma F, Zeng XJ, Wang J. A knowledge-free path planning approach for smart ships based on reinforcement learning. Ocean Engineering, 2019, 189: 106299

[8] Chen X, Ghadirzadeh A, Folkesson J, Björkman M, Jensfelt P (2018) Deep reinforcement learning to acquire navigation skills for wheel-legged robots in complex environments. In 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 3110–3116. https://doi.org/10.48550/arXiv.1804.10500

[9] Chen X, Yang Y, Wang S, Wu H, Tang J, Zhao J, Wang Z. Ship type recognition via a coarse-to-fine cascaded convolution neural network. Journal of Navigation, 2020, 73(4): 813-832

[10] Chen Z, Chen D, Zhang Y, Cheng X, Zhang M, Wu C. Deep learning for autonomous ship-oriented small ship detection. Safety Science, 2020, 130: 104812

[11] Cheng Y, Zhang W. Concise deep reinforcement learning obstacle avoidance for underactuated unmanned marine ships. Neurocomputing, 2018, 272(1): 63-73

[12] Crisp DJ (2004) The state-of-the-art in ship detection in synthetic aperture radar imagery. organic letters. https://doi.org/10.1002/chin.200442210

[13] Dahl GE, Yu D, Deng L, Acero A. Context-dependent pre-trained deep neural networks for large-vocabulary speech recognition. IEEE Transactions on audio, speech, and language processing, 2011, 20(1): 30-42

[14] Dauphin Y, Pascanu R, Gulcehre C, Cho K, Ganguli S, Bengio Y (2014) Identifying and attacking the saddle point problem in high-dimensional non-convex optimization. arXiv preprint arXiv: 1406.2572

[15] Davarzani H, Fahimnia B, Bell M, Sarkis J. Greening ports and maritime logistics: A review. Transportation Research Part D: Transport and Environment, 2016, 48: 473-487

[16] Desai N. Dynamic positioning: Method for disaster prevention and risk management. Procedia Earth and Planetary Science, 2015, 11: 216-223 Global Challenges, Policy Framework & Sustainable Development for Mining of Mineral and Fossil Energy Resources (GCPF: 2015–20)

[17] Do K. Practical formation control of multiple underactuated ships with limited sensing ranges. Robotics and Autonomous Systems, 2011, 59(6): 457-471

[18] Fan Q, Chen F, Cheng M, Lou S, Li J. Ship detection using a fully convolutional network with compact polari-metric SAR images. Remote Sensing, 2019, 11(18): 2171

[19] Farahnakian F, Heikkonen J. Deep learning based multi-modal fusion architectures for maritime vessel detection. Remote Sensing, 2020, 12(16): 2509

[20] Fu K, Li Y, Sun H, Yang X, Xu G, Li Y, Sun X. A ship rotation detection model in remote sensing images based on feature fusion pyramid network and deep reinforcement learning. Remote Sensing, 2018, 10(12): 1922

[21] Gallego A, Pertusa A, Gil P. Automatic ship classification from optical aerial images with convolutional neural networks. Remote Sensing, 2018, 10(4): 511

[22] Gao F, He Y, Wang J, Hussain A, Zhou H. Anchor-free convolutional network with dense attention feature aggregation for ship detection in sar images. Remote Sensing, 2020, 12(16): 2619

[23] Gao F, Shi W, Wang J, Yang E, Zhou H. Enhanced feature extraction for ship detection from multi- resolution and multi-scene synthetic aperture radar (SAR) images. Remote Sensing, 2019, 11(22): 2694

[24] Gao L, He Y, Sun X, Jia X, Zhang B. Incorporating negative sample training for ship detection based on deep learning. Sensors, 2019, 19(3): 684

[25] García RF, Vázquez CSC. Introduction to ship dynamic positioning systems. Journal of Maritime Research, 2008, 5(1): 79-95

[26] Gu Y, Goez JC, Guajardo M, Wallace SW. Autonomous ships: state of the art and potential opportunities in logistics. International Transactions in Operational Research, 2020, 28: 1706-1739

[27] Guo M, Guo C, Zhang C, Zhang D, Gao Z. Fusion of ship perceptual information for electronic navigational chart and radar images based on deep learning. Journal of Navigation, 2020, 73(1): 192-211

[28] Guo S, Zhang X, Zheng Y, Du Y. An autonomous path planning model for unmanned ships based on deep reinforcement learning. Sensors, 2020, 20(2): 426

[29] Han Z, Chong J (2004) A review of ship detection algorithms in polarimetric sar images. In International Conference on Signal Processing. https://doi.org/10.1109/ICOSR2004.1442203

[30] Hass FS, Arsanjani JJ. Deep learning for detecting and classifying ocean objects: Application of yolov3 for iceberg-ship discrimination. International Journal of Geo-Information, 2020, 9(12): 758

[31] Heij C, Knapp S. Predictive power of inspection outcomes for future shipping accidents-an empirical appraisal with special attention for human factor aspects. Maritime Policy & Management, 2018, 45(5–6): 604-621

[32] Heiselberg H. Ship-iceberg classification in sar and multispectral satellite images with neural networks. Remote Sensing, 2020, 12(15): 2353

[33] Hinton GE. Deep belief networks. Scholarpedia, 2009, 4(5): 5947

[34] Howard IM (1994) A review of rolling element bearing vibration detection, diagnosis and prognosis. NASA STI/Recon Technical Report N

[35] Huang B, He B, Wu L, Lin Y. A deep learning approach to detecting ships from high-resolution aerial remote sensing images. Journal of Coastal Research, 2020, 111(sp1): 16-20

[36] Huang G, Wan Z, Liu X, Hui J, Wang Z, Zhang Z. Ship detection based on squeeze excitation skip-connection path networks for optical remote sensing images. Neurocomputing, 2019, 332: 215-223

[37] IMO (2021) Imo takes first steps to address autonomous ships. http://www.imo.org/en/mediacentre/pressbriefings/pages/08-msc-99-mass-scoping.aspx. [Accessed May 4, 2021]

[38] IMO (2019) Interim guidelines for mass trials. https://wwwcdn.imo.org/localresources/en/MediaCentre/HotTopics/Documents/MSC.1-Circ.1604%20-%20Interim%20Guidelines%20For%20Mass%20Trials%20(Secretariat).pdf

[39] Inggs M, Robinson A (1995) Neural approaches to ship target recognition. In Proceedings International Radar Conference, 386–391. https://doi.org/10.1109/RADAR.1995.522577

[40] Jiang B, Ma X, Lu Y, Li Y, Feng L, Shi Z. Ship detection in spaceborne infrared images based on convolutional neural networks and synthetic targets. Infrared Physics & Technology, 2019, 97: 229-234

[41] Jiao J, Zhang Y, Sun H, Yang X, Gao X, Hong W, Fu K, Sun X. A densely connected end-to-end neural network for multiscale and multiscene sar ship detection. IEEE Access, 2018, 6: 20881-20892

[42] Jo K, Kim J, Kim D, Jang C, Sunwoo M. Development of autonomous car-part ii: A case study on the implementation of an autonomous driving system based on distributed architecture. IEEE Transactions on Industrial Electronics, 2015, 62(8): 5119-5132

[43] Jose MGS, Juan FJ. State-of-the-art of wave measurement for ship motion prediction. IFAC Proceedings Volumes, 2010, 43(20): 295-300 8th IFAC Conference on Control Applications in Marine Systems

[44] Joseph SIT, Sasikala J, Sujitha Juliet D. A novel ship detection and classification algorithm using a deep learning neural network model with morphological processing (M-DLNN). Soft Computing, 2018, 23: 2693-2700

[45] Jurgen and Ronald K (2013) Autonomous vehicles for safer driving. SAE International

[46] Kang M, Leng X, Lin Z, Ji K (2017) A modified faster r-cnn based on cfar algorithm for sar ship detection. In 2017 International Workshop on Remote Sensing with Intelligent Processing (RSIP), 1–4. https://doi.org/10.1109/RSIP.2017.7958815

[47] Kim H, Kim D, Park B, Lee SM. Artificial intelligence vision-based monitoring system for ship berthing. IEEE Access, 2020, 8: 227014-227023

[48] Kim H, Koo J, Kim D, Park B, Jo Y, Myung H, Lee D. Vision-based real-time obstacle segmentation algorithm for autonomous surface vehicle. IEEE Access, 2019, 7: 179420-179428

[49] Kim KI, Lee KM. Deep learning-based caution area traffic prediction with automatic identification system sensor data. Sensors, 2018, 18(9): 3172

[50] Krizhevsky A, Sutskever I, Hinton GE. Imagenet classification with deep convolutional neural networks, 2012, Red Hook, NY, USA: Curran Associates Inc., 1097-1105

[51] Lazarowska A (2019) Research on algorithms for autonomous navigation of ships. WMU Journal of Maritime Affairs, 18. https://doi.org/10.1007/s13437-019-00172-0

[52] LeCun Y, Boser B, Denker JS, Henderson D, Howard RE, Hubbard W, Jackel LD. Backpropagation applied to handwritten zip code recognition. Neural Computation, 1989, 1(4): 541-551

[53] Lee D, Lee SJ, Yim SC. Reinforcement learning-based adaptive pid controller for dps. Ocean Engineering, 2020, 216: 108053

[54] Li J, Qu C, Peng S. Ship classification for unbalanced sar dataset based on convolutional neural network. Journal of Applied Remote Sensing, 2018, 12(03): 035010

[55] Li S, Zhang Z, Li B, Li C. Multiscale rotated bounding box-based deep learning method for detecting ship targets in remote sensing images. Sensors, 2018, 18(8): 2702

[56] Li Z, Zhao L, Han X, Pan M, Hwang FJ. Lightweight ship detection methods based on yolov3 and densenet. Mathematical Problems in Engineering, 2020, 2020: 1-10

[57] Lin S, Feng C, Chen Z. A ship target location and mask generation algorithms base on mask rcnn. International Journal of Computational Intelligence Systems, 2019, 12(2): 1134-1143

[58] Lin Z, Ji K, Leng X, Kuang G. Squeeze and excitation rank faster r-cnn for ship detection in sar images. IEEE Geoscience and Remote Sensing Letters, 2019, 16(5): 751-755

[59] Liu B, Wang SZ, Xie ZX, Zhao JS, Li MF. Ship recognition and tracking system for intelligent ship based on deep learning framework. TransNav the International Journal on Marine Navigation and Safety of Sea Transportation, 2019, 13(4): 699-705

[60] Liu Y, Duan W, Huang L, Duan S, Ma X. The input vector space optimization for lstm deep learning model in real-time prediction of ship motions. Ocean Engineering, 2020, 213: 107681

[61] Lloyd’s Register (2015) Design code for unmanned marine systems. https://www.cdinfo.lr.org/information/documents/ShipRight/Design%20and%20Construction/Additional%20Design%20Procedures/Design%20Code%20for%20Unmanned%20Marine%20Systems/Design%20Code%20for%20Unmanned%20Marine%20Systems,%20February%202017.pdf. [Accessed February 4, 2021]

[62] Lv Y, Duan Y, Kang W, Li Z, Wang FY. Traffic flow prediction with big data: A deep learning approach. IEEE Transactions on Intelligent Transportation Systems, 2015, 16(2): 865-873

[63] McCulloch WS, Pitts W. A logical calculus of the ideas immanent in nervous activity. Bulletin of Mathematical Biophysics, 1943, 5: 115-133

[64] Minsky ML. Computation: Finite and infinite machines, 1968, USA: Prentice-Hall, Inc.

[65] Morillas JRA, García IC, Zölzer U (2015) Ship detection based on svm using color and texture features. In 2015 IEEE International Conference on Intelligent Computer Communication and Processing (ICCP), 343–350. https://doi.org/10.1109/ICCP.2015.7312682

[66] Munim ZH. Autonomous ships: a review, innovative applications and future maritime business models. Supply Chain Forum, 2019, 20(4): 266-279

[67] MUNIN (2021)About munin-maritime unmanned navigation through intelligence in networks. http://www.unmanned-ship.org/munin/about/ Accessed February 4, 2021

[68] Nair V, Hinton G. Rectified linear units improve restricted boltzmann machines vinod nair. Proceedings of ICML, 2010, 27: 807-814

[69] Niu H, Ozanich E, Gerstoft P. Ship localization in santa barbara channel using machine learning classifiers. Journal of the Acoustical Society of America, 2017, 142(5): EL455-EL460

[70] Noel A, Shreyanka K, Kumar K, Shameem BM, Akshar B (2019) Autonomous ship navigation methods: A review. In International Conference on Marine Engineering and Technology Oman. https://doi.org/10.24868/icmet.oman.2019.028

[71] Ozturk U, Cicek K. Individual collision risk assessment in ship navigation: A systematic literature review. Ocean Engineering, 2019, 180: 130-143

[72] Perera LP (2018) Autonomous ship navigation under deep learning and the challenges in colregs. In International Conference on Offshore Mechanics and Arctic Engineering, volume Volume 11B: Honoring Symposium for Professor Carlos Guedes Soares on Marine Technology and Ocean Engineering. https://doi.org/10.1115/OMAE2018-77672

[73] Phan N, Dou D, Wang H, Kil D, Piniewski B. Ontology-based deep learning for human behavior prediction with explanations in health social networks. Information Sciences, 2017, 384: 298-313

[74] Rane A, Sangili V. Implementation of improved ship-iceberg classifier using deep learning. Journal of Intelligent Systems, 2019, 29(1): 1514-1522

[75] Roberts G. Trends in marine control systems. IFAC Proceedings Volumes, 2007, 40(17): 315-322

[76] Rosenblatt F. The perceptron: a probabilistic model for information storage and organization in the brain. Psychological Review, 1958, 65: 386-408

[77] Rosenblatt F (1962) Principles of neurodynamics: Perceptrons and the theory of brain mechanisms. Cornell Aeronautical Laboratory. https://doi.org/10.1001/archpsyc.1962.01720030064010

[78] Rumelhart DE, Hinton GE, Williams RJ. Learning representations by back propagating errors. Nature, 1986, 323(6088): 533-536

[79] Sallab AE, Abdou M, Perot E, Yogamani S. Deep reinforcement learning framework for autonomous driving. Electronic Imaging, 2017, 2017(19): 70-76

[80] Schroder-Hinrichs J. Human and organizational factors in the maritime world - are we keeping up to speed? WMU Journal of Maritime Affairs, 2010, 9(1): 1-3

[81] Schwegmann CP, Kleynhans W, Salmon BP, Mdakane LW, and Meyer RGV (2016). Very deep learning for ship discrimination in synthetic aperture radar imagery. In 2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), 104–107. https://doi.org/10.1109/IGARSS.2016.7729017

[82] Senior A, Vanhoucke V, Nguyen P, Sainath T, et al. (2012). Deep neural networks for acoustic modeling in speech recognition. IEEE Signal processing magazine, https://doi.org/10.1109/MSP.2012.2205597

[83] Shao J, Qu L, Li J, Peng S. A lightweight convolutional neural network based on visual attention for sar image target classification. Sensors, 2018, 18(9): 3039

[84] Sharifzadeh F, Akbarizadeh G, Kavian YS. Ship classification in sar images using a new hybrid cnn-mlp classifier. Journal of The Indian Society of Remote Sensing, 2019, 47(4): 551-562

[85] Shen S, Yang H, Li J, Xu G, Sheng M. Auditory inspired convolutional neural networks for ship type classification with raw hydrophone data. Entropy, 2018, 20(12): 990

[86] Shen S, Yang H, Yao X, Li J, Xu G, Sheng M. Ship type classification by convolutional neural networks with auditory-like mechanisms. Sensors, 2020, 20(1): 253

[87] Shi Q, Li W, Zhang F, Hu W, Sun X, Gao L. Deep cnn with multi-scale rotation invariance features for ship classification. IEEE Access, 2018, 6: 38656-38668

[88] Shi Y, Shen C, Fang H, Li H. Advanced control in marine mechatronic systems: A survey. IEEE/ASME Transactions on Mechatronics, 2017, 22(3): 1121-1131

[89] Silver D, Schrittwieser J, Simonyan K, Antonoglou I, Hassabis D. Mastering the game of go without human knowledge. Nature, 2017, 550(7676): 354-359

[90] Statheros T, Howells G, McDonald-Maier K. Autonomous ship collision avoidance navigation concepts, technologies and techniques. Journal of Navigation, 2008, 61(1): 129-142

[91] Sun W, Gao X (2021) Deep learning-based trajectory tracking control for unmanned surface vehicle. Mathematical Problems in Engineering, 2021. https://doi.org/10.1155/2021/8926738

[92] Tuyen LP, Layek A, Vien NA, Chung T (2017) Deep reinforcement learning algorithms for steering an underactuated ship. In 2017 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI), 602–607. https://doi.org/10.1109/MFI.2017.8170388

[93] Vandeginste B, Massart D, Buydens L, De Jong S, Lewi P, Smeyers-Verbeke J (1998). Chapter 44-artificial neural networks. In Vandeginste B, Massart D, Buydens L, De Jong S, Lewi P, Smeyers-Verbeke J, edited. Handbook of Chemometrics and Qualimetrics: Part B, volume 20 of Data Handling in Science and Technology, 649–699. Elsevier

[94] Vinyals O, Babuschkin I, Czarnecki WM, Mathieu M, Dudzik A, Chung J, Choi DH, Powell R, Ewalds T, Georgie P. Grandmaster level in starcraft ii using multi- agent reinforcement learning. Nature, 2019, 575: 350-354

[95] Wang J, Lu C, Jiang W. Simultaneous ship detection and orientation estimation in sar images based on attention module and angle regression. Sensors, 2018, 18(9): 2851

[96] Wang R, Li J, Duan Y, Cao H, Zhao Y. Study on the combined application of cfar and deep learning in ship detection. Journal of the Indian Society of Remote Sensing, 2018, 46: 1413-1421

[97] Wang X, Jiang R, Li L, Lin Y, Zheng X, Wang FY. Capturing car-following behaviors by deep learning. IEEE Transactions on Intelligent Transportation Systems, 2018, 19(3): 910-920

[98] Wang Y, Tong J, Song TY, Wan ZH (2018b) Unmanned surface vehicle course tracking control based on neural network and deep deterministic policy gradient algorithm. In 2018 OCEANS-MTS/IEEE Kobe Techno-Oceans (OTO), 1–5. https://doi.org/10.1109/OCEANSKOBE.2018.8559329

[99] Wang Y, Wang C, Zhang H. Ship classification in highresolution sar images using deep learning of small datasets. Sensors, 2018, 18(9): 2929

[100] Wang Y, Wang C, Zhang H, Dong Y, Wei S (2019a) Automatic ship detection based on retinanet using multi-resolution gaofen-3 imagery. Remote Sensing, 11(5). https://doi.org/10.3390/rs11050531

[101] Wang Y, Wang C, Zhang H, Dong Y, Wei S. A sar dataset of ship detection for deep learning under complex backgrounds. Remote Sensing, 2019, 11(7): 765

[102] Wang Z, Zhou Y, Wang F, Wang S, Xu Z. Sdgh-net: Ship detection in optical remote sensing images based on Gaussian heatmap regression. Remote Sensing, 2021, 13(3): 499

[103] Weichselbaum J, Zinner C, Gebauer O, Pree W. Accurate 3d-vision-based obstacle detection for an autonomous train. Computers in Industry, Special Issue: 3D Imaging in Industry, 2013, 64(9): 1209-1220

[104] Werbos PJ. Generalization of backpropagation with application to a recurrent gas market model. Neural Networks, 1988, 1(4): 339-356

[105] Wu DF, Yang GH (2014) Review on key techniques for ship dynamic positioning system. Ship Science and Technology

[106] Ye J, Roy S, Godjevac M, Reppa V, Baldi S. Robustifying dynamic positioning of crane vessels for heavy lifting operation. IEEE/CAA Journal of Automatica Sinica, 2021, 8(4): 753-765

[107] Yin J, Tang T, Yang L, Xun J, Huang Y, Gao Z. Research and development of automatic train operation for railway transportation systems: A survey. Transportation Research Part C: Emerging Technologies, 2017, 85: 548-572

[108] You Y, Li Z, Ran B, Cao J, Liu F. Broad area target search system for ship detection via deep convolutional neural network. Remote Sensing, 2019, 11(17): 1965

[109] Zhang D, Zhan J, Tan L, Gao Y, Upan R. Comparison of two deep learning methods for ship target recognition with optical remotely sensed data. Neural Computing and Applications, 2020, 33: 4639-4649

[110] Zhang Q, Im N-K, Ding Z, Zhang M. Su SF, Wang N. Review on the Research of Ship Automatic Berthing Control. Offshore Robotics, 2022, Singapore: Springer

[111] Zhang Q, Pan W, Reppa V (2020b) Model-reference reinforcement learning control of autonomous surface vehicles. In 2020 59th IEEE Conference on Decision and Control (CDC), pages 5291–5296, https://doi.org/10.1109/CDC42340.2020.9304347

[112] Zhang Q, Zhu G, Hu X, Yang R. Adaptive neural network auto-berthing control of marine ships. Ocean Engineering, 2019, 177: 40-48

[113] Zhang T, Zhang X. High-speed ship detection in sar images based on a grid convolutional neural network. Remote Sensing, 2019, 11(10): 1206

[114] Zhang T, Zhang X, Shi J, Wei S. Depthwise separable convolution neural network for high-speed sar ship detection. Remote Sensing, 2019, 11(21): 2483

[115] Zhao H, Zhang W, Sun H, Xue B. Embedded deep learning for ship detection and recognition. Future Internet, 2019, 11(2): 53

[116] Zhao L, Roh MI. Colregs-compliant multiship collision avoidance based on deep reinforcement learning. Ocean Engineering, 2019, 191: 106436

[117] Zhao S, Xu Y, Li W, Lang H. Optical remote sensing ship image classification based on deep feature combined distance metric learning. Journal of Coastal Research, 2020, 102(sp1): 82-87

[118] Zhong C, Jiang Z, Chu X, Liu L. Inland ship trajectory restoration by recurrent neural network. Journal of Navigation, 2019, 72(6): 1359-1377

[119] Zhu M, Wang X, Wang Y. Human-like autonomous car-following model with deep reinforcement learning. Transportation Research Part C: Emerging Technologies, 2018, 97: 348-368

[120] Zou L, Zhang H, Wang C, Wu F, Gu F. Mw-acgan: Generating multiscale high-resolution sar images for ship detection. Sensors, 2020, 20(22): 6673