Incremental Learning Based on Anchored Class Centers for SAR Automatic Target Recognition

被引:20
|
作者
Li, Bin [1 ]
Cui, Zongyong [1 ]
Cao, Zongjie [1 ]
Yang, Jianyu [1 ]
机构
[1] Univ Elect Sci & Technol China, Sch Informat & Commun Engn, Chengdu 611731, Peoples R China
来源
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING | 2022年 / 60卷
基金
中国国家自然科学基金;
关键词
Synthetic aperture radar; Measurement; Training; Target recognition; Azimuth; Computational modeling; Adaptation models; Catastrophic forgetting; class anchor clustering (CAC); incremental learning; synthetic aperture radar automatic target recognition (SAR ATR);
D O I
10.1109/TGRS.2022.3208346
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
Although deep learning methods have achieved great success in synthetic aperture radar automatic target recognition (SAR ATR), their accuracies decline sharply, as new classes are learned, which is known as catastrophic forgetting. The overlapping or confusion between the representations of new and old classes in the feature space is the main cause of catastrophic forgetting. In this article, the incremental class anchor clustering (ICAC) is proposed to address this issue. ICAC solves this problem from three perspectives: 1) how to learn the new classes; 2) how to enable the model to recognize and classify the old classes; and 3) how to solve the imbalance between old classes and new classes. To learn the new classes, ICAC adaptively adds new anchored class centers for new classes, and the features of each new class will be clustered around the corresponding anchored class center. To enable the model to recognize and classify the old classes, ICAC stores some exemplars for the old classes to ensure the classification ability of the old classes without losing the old class centers in the feature space. At the same time, ICAC adopts knowledge distillation to further alleviate catastrophic forgetting. To solve the imbalance between old classes and new classes, ICAC proposes a learning strategy named separable learning (SL), which computes the losses of the old and new exemplars separately and then adds the two losses to make a gradient descent. Experiments on the moving and stationary target acquisition and recognition (MSTAR) dataset and the OpenSARShip dataset demonstrate the effectiveness of this method in SAR ATR.
引用
收藏
页数:13
相关论文
共 50 条
  • [1] Incremental SAR Automatic Target Recognition With Error Correction and High Plasticity
    Tang, Jiaxin
    Xiang, Deliang
    Zhang, Fan
    Ma, Fei
    Zhou, Yongsheng
    Li, HengChao
    IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2022, 15 : 1327 - 1339
  • [2] Dynamic Embedding Relation Distillation Network for Incremental SAR Automatic Target Recognition
    Ren, Haohao
    Dong, Fulu
    Zhou, Rongsheng
    Yu, Xuelian
    Zou, Lin
    Zhou, Yun
    IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2024, 21 : 1 - 5
  • [3] SAR Incremental Automatic Target Recognition Based on Mutual Information Maximization
    Li, Bin
    Cui, Zongyong
    Wang, Haohan
    Deng, Yijie
    Ma, Jizhen
    Yang, Jianyu
    Cao, Zongjie
    IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2024, 21 : 1 - 5
  • [4] Class Boundary Exemplar Selection Based Incremental Learning for Automatic Target Recognition
    Dang, Sihang
    Cao, Zongjie
    Cui, Zongyong
    Pi, Yiming
    Liu, Nengyuan
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2020, 58 (08): : 5782 - 5792
  • [5] Distribution Reliability Assessment-Based Incremental Learning for Automatic Target Recognition
    Dang, Sihang
    Cui, Zongyong
    Cao, Zongjie
    Pi, Yiming
    Feng, Xiaoyi
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2023, 61
  • [6] IncSAR: A Dual Fusion Incremental Learning Framework for SAR Target Recognition
    Karantaidis, George
    Pantsios, Athanasios
    Kompatsiaris, Ioannis
    Papadopoulos, Symeon
    IEEE ACCESS, 2025, 13 : 12358 - 12372
  • [7] SAR Target Incremental Recognition Based on Features With Strong Separability
    Gao, Fei
    Kong, Lingzhe
    Lang, Rongling
    Sun, Jinping
    Wang, Jun
    Hussain, Amir
    Zhou, Huiyu
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2024, 62 : 1 - 13
  • [8] Robust SAR Automatic Target Recognition Via Adversarial Learning
    Guo, Yuchen
    Du, Lan
    Wei, Di
    Li, Chen
    IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2021, 14 (14) : 716 - 729
  • [9] Few-Shot Class-Incremental SAR Target Recognition via Orthogonal Distributed Features
    Kong, Lingzhe
    Gao, Fei
    He, Xiaoyu
    Wang, Jun
    Sun, Jinping
    Zhou, Huiyu
    Hussain, Amir
    IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS, 2025, 61 (01) : 325 - 341
  • [10] Multilevel Adaptive Knowledge Distillation Network for Incremental SAR Target Recognition
    Yu, Xuelian
    Dong, Fulu
    Ren, Haohao
    Zhang, Chengfa
    Zou, Lin
    Zhou, Yun
    IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2023, 20