• Home
  • Search
  • Semi-Supervised Incremental Learning Framework with Uncertainty Assessment Mechanism for Radar Target Recognition
  • https://doi.org/10.1109/iccc68654.2025.11437912Copy DOI Icon

Semi-Supervised Incremental Learning Framework with Uncertainty Assessment Mechanism for Radar Target Recognition

  • Dec 12, 2025
  • Lu Qian +5 more
Show More
  • Abstract
  • Literature Map
  • References
  • Similar Papers
Abstract

In realistic radar application scenarios, numerous unannotated samples of new categories can be acquired continuously. It is very meaningful to make full use of these unannotated samples so that the recognition model can continuously generalize to new categories at a lower annotation cost. To this end, this letter proposes a semi-supervised incremental learning framework with uncertainty assessment named high-confidence pseudolabeling network (HPN) for radar target recognition tasks. Specifically, to update the recognition model with a large number of unannotated samples, an adaptive uncertainty measurement method is first proposed to calculate the confidence of unknown samples. Then, a confidence assessment mechanism based on Gaussian distribution maximization is presented to evaluate the correctness of pseudo-labeling, aiming to eliminate unreliable pseudo-labeling samples with high confidence. Next, the herdingbased clustering method is exploited to select representative exemplars of each old target, so as to recall the knowledge of old categories during incremental learning. Finally, a multi-task incremental learning loss is presented, which enables the model to continuously identify new categories of targets in a semisupervised learning framework. Evaluation experiments on the two datasets, i.e., moving and stationary target acquisition and recognition (MSTAR) and synthesis and measurement of pairs of labels experiment (SAMPLE), validate the effectiveness of the proposed method for incremental radar target recognition.

Similar Papers
  • Research Article

Multivariate empirical mode decomposition with application to SAR image target recognition

  • Jan 01, 2021
  • Infrared and Laser Engineering
  • 伍友龙 Wu Youlong
  • Research Article

Unsupervised Feature Selection-Driven Active Learning for Semi-Supervised Automatic ECG Analysis.

  • Jan 01, 2025
  • IEEE journal of biomedical and health informatics
  • Xiao Li +8
  • Research Article
  • Citations24

Multilevel Adaptive Knowledge Distillation Network for Incremental SAR Target Recognition

  • Jan 01, 2023
  • IEEE Geoscience and Remote Sensing Letters
  • Xuelian Yu +5
  • Conference Article
  • Citations2

Automatic target recognition method for low-resolution ground surveillance radar based on 1D-CNN

  • Jan 20, 2021
  • Renhong Xie +5
  • Research Article
  • Citations8

Dynamic Embedding Relation Distillation Network for Incremental SAR Automatic Target Recognition

  • Jan 01, 2024
  • IEEE Geoscience and Remote Sensing Letters
  • Haohao Ren +5
  • PDF
  • Research Article
  • Citations23

Target Recognition in SAR Images Based on Information-Decoupled Representation

  • Jan 19, 2018
  • Remote Sensing
  • Ming Chang +1
  • Research Article
  • Citations46

Cross-task and cross-domain SAR target recognition: A meta-transfer learning approach

  • Feb 07, 2023
  • Pattern Recognition
  • Yukun Zhang +3
  • PDF
  • Research Article
  • Citations6

SAR Target Recognition via Joint Sparse and Dense Representation of Monogenic Signal

  • Nov 15, 2019
  • Remote Sensing
  • Meiting Yu +3
  • Conference Article
  • Citations1

Research on Optimization Algorithm of Low-Resolution Radar Target Recognition Based on Neural Network

  • Nov 01, 2022
  • Liu Xiaoyan +7
  • Conference Article
  • Citations2

A multi-scale temporal model for multi-view radar high-resolution range profile recognition

  • Dec 01, 2017
  • Xuan Peng +3
  • Research Article
  • Citations21

A Novel Semi-Supervised Long-Tailed Learning Framework With Spatial Neighborhood Information for Hyperspectral Image Classification

  • Jan 01, 2023
  • IEEE Geoscience and Remote Sensing Letters
  • Yining Feng +4
  • Research Article
  • Citations16

Combination of global and local filters for robust SAR target recognition under various extended operating conditions

  • Oct 05, 2018
  • Information Sciences
  • Baiyuan Ding +1
  • Supplementary Content

Transformers and contrastive semi-supervised learning for medical image segmentation

  • Jan 01, 2026
  • Enlighten: Theses (The University of Glasgow)
  • Qianying Liu
  • Research Article
  • Citations35

Developing a semi-supervised learning and ordinal classification framework for quality level prediction in manufacturing

  • May 09, 2023
  • Computers & Industrial Engineering
  • Gyeongho Kim +3
  • Conference Article
  • Citations5

Multi-View Fusion Based on Expectation Maximization for SAR Target Recognition

  • Sep 26, 2020
  • Yukun Zhang +4
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.