A Bayesian Approach to Active Self-Paced Deep Learning for SAR Automatic Target Recognition

被引:18
作者
Ren, Haohao [1 ]
Yu, Xuelian [1 ]
Bruzzone, Lorenzo [2 ]
Zhang, Yukun [1 ]
Zou, Lin [1 ]
Wang, Xuegang [1 ]
机构
[1] Univ Elect Sci & Technol China, Sch Informat & Commun Engn, Chengdu 611731, Peoples R China
[2] Univ Trento, Dept Informat Engn & Comp Sci, I-38123 Trento, Italy
基金
中国国家自然科学基金;
关键词
Synthetic aperture radar; Radar polarimetry; Bayes methods; Labeling; Data models; Neural networks; Target recognition; Active learning (AL); automatic target recognition (ATR); Bayesian learning; convolutional neural network (CNN); remote sensing; synthetic aperture radar (SAR); ATR;
D O I
10.1109/LGRS.2020.3036585
中图分类号
P3 [地球物理学]; P59 [地球化学];
学科分类号
0708 ; 070902 ;
摘要
Deep learning has attracted intensive attention in synthetic aperture radar (SAR) automatic target recognition (ATR). Usually, a considerable number of labeled samples are necessary to learn a deep model for obtaining good generalization capability. However, the process of sample labeling is time-consuming and costly. This letter proposes an active self-paced deep learning (ASPDL) approach to SAR ATR. In a nutshell, we first introduce the Bayesian inference into the process of deep model parameter optimization, aiming at learning a robust classification model in the case of a limited number of labeled samples. Next, a cost-effective sample selection strategy is presented to iteratively and actively select the informative samples from a pool of unlabeled samples for labeling. Concretely, high-confidence samples are actively selected through self-paced learning (SPL) way and automatically pseudo-labeled with the current classification model, whereas low-confidence samples are chosen through an active learning strategy and manually labeled. Finally, we update the parameters of the model by minimizing a dual-loss function using a new training set that is constructed by incorporating new labeled samples with original ones. Experiments on the moving and stationary target acquisition and recognition (MSTAR) benchmark data demonstrate that the proposed method can achieve better classification accuracy with relatively few labeled samples compared with some state-of-the-art methods.
引用
收藏
页数:5
相关论文
共 50 条
[41]   SAR Target Recognition Based on Cross-Domain and Cross-Task Transfer Learning [J].
Wang, Ke ;
Zhang, Gong ;
Leung, Henry .
IEEE ACCESS, 2019, 7 :153391-153399
[42]   What, Where, and How to Transfer in SAR Target Recognition Based on Deep CNNs [J].
Huang, Zhongling ;
Pan, Zongxu ;
Lei, Bin .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2020, 58 (04) :2324-2336
[43]   Semisupervised Hotspot Detection With Self-Paced Multitask Learning [J].
Chen, Ying ;
Lin, Yibo ;
Gai, Tianyang ;
Su, Yajuan ;
Wei, Yayi ;
Pan, David Z. .
IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS, 2020, 39 (07) :1511-1523
[44]   SAR-to-Virtual Optical Image Translation for Improving SAR Automatic Target Recognition [J].
Lee, In Ho ;
Park, Chan Gook .
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2023, 20
[45]   Hierarchical Distribution-Based Exemplar Replay for Incremental SAR Automatic Target Recognition [J].
Ren, Haohao ;
Zhou, Rongsheng ;
Zou, Lin ;
Tang, Hao .
IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS, 2025, 61 (03) :6576-6588
[46]   SAR Automatic Target Recognition Based on Deep Convolutional Neural Network [J].
Xu, Ying ;
Liu, Kaipin ;
Ying, Zilu ;
Shang, Lijuan ;
Liu, Jian ;
Zhai, Yikui ;
Piuri, Vincenzo ;
Scotti, Fabio .
IMAGE AND GRAPHICS (ICIG 2017), PT III, 2017, 10668 :656-667
[47]   L2,1-Constrained Deep Incremental NMF Approach for SAR Automatic Target Recognition [J].
Cao, Changjie ;
Chou, Ran ;
Zhang, Hanyue ;
Li, Xian ;
Luo, Tangyun ;
Liu, Bingli .
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2024, 21 :1-5
[48]   Occluded Target Recognition in SAR Imagery With Scattering Excitation Learning and Channel Dropout [J].
He, Dunyun ;
Guo, Weiwei ;
Zhang, Tao ;
Zhang, Zenghui ;
Yu, Wenxian .
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2023, 20
[49]   A neural system for automatic target learning and recognition applied to bare and camouflaged SAR targets [J].
Bernardon, AM ;
Carrick, JE .
NEURAL NETWORKS, 1995, 8 (7-8) :1103-1108
[50]   Pyramidal Dilation Attention Convolutional Network With Active and Self-Paced Learning for Hyperspectral Image Classification [J].
Hou, Wenhui ;
Chen, Na ;
Peng, Jiangtao ;
Sun, Weiwei ;
Du, Qian .
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2023, 16 :1503-1518