IncSAR: A Dual Fusion Incremental Learning Framework for SAR Target Recognition
Deep learning techniques have achieved significant success in Synthetic Aperture Radar (SAR) target recognition using predefined datasets in static scenarios. However, real-world applications demand that models incrementally learn new information without forgetting previously acquired knowledge. The...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10838563/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832590320646225920 |
|---|---|
| author | George Karantaidis Athanasios Pantsios Ioannis Kompatsiaris Symeon Papadopoulos |
| author_facet | George Karantaidis Athanasios Pantsios Ioannis Kompatsiaris Symeon Papadopoulos |
| author_sort | George Karantaidis |
| collection | DOAJ |
| description | Deep learning techniques have achieved significant success in Synthetic Aperture Radar (SAR) target recognition using predefined datasets in static scenarios. However, real-world applications demand that models incrementally learn new information without forgetting previously acquired knowledge. The challenge of catastrophic forgetting, where models lose past knowledge when adapting to new tasks, remains a critical issue. In this paper, we introduce IncSAR, an incremental learning framework designed to tackle catastrophic forgetting in SAR target recognition. IncSAR combines the power of a Vision Transformer (ViT) and a custom-designed Convolutional Neural Network (CNN) in a dual-branch architecture, integrated via a late-fusion strategy. Additionally, we explore the use of TinyViT to reduce computational complexity and propose an attention mechanism to dynamically enhance feature representation. To mitigate the speckle noise inherent in SAR images, we employ a denoising module based on a neural network approximation of Robust Principal Component Analysis (RPCA), leveraging a simple neural network for efficient noise reduction in SAR imagery. Moreover, a random projection layer improves the linear separability of features, and a variant of Linear Discriminant Analysis (LDA) decorrelates extracted class prototypes for better generalization. Extensive experiments on the MSTAR, SAR-AIRcraft-1.0, and OpenSARShip benchmark datasets demonstrate that IncSAR significantly outperforms state-of-the-art approaches, achieving a 99.63% average accuracy and a 0.33% performance drop, representing an 89% improvement in retention compared to existing techniques. The source code is available at <uri>https://github.com/geokarant/IncSAR</uri>. |
| format | Article |
| id | doaj-art-427b3e985c1049d4a064c700c28ea3c3 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-427b3e985c1049d4a064c700c28ea3c32025-01-24T00:01:39ZengIEEEIEEE Access2169-35362025-01-0113123581237210.1109/ACCESS.2025.352863310838563IncSAR: A Dual Fusion Incremental Learning Framework for SAR Target RecognitionGeorge Karantaidis0https://orcid.org/0000-0002-7980-7805Athanasios Pantsios1Ioannis Kompatsiaris2https://orcid.org/0000-0001-6447-9020Symeon Papadopoulos3https://orcid.org/0000-0002-5441-7341Centre for Research and Technology Hellas, Thessaloniki, GreeceCentre for Research and Technology Hellas, Thessaloniki, GreeceCentre for Research and Technology Hellas, Thessaloniki, GreeceCentre for Research and Technology Hellas, Thessaloniki, GreeceDeep learning techniques have achieved significant success in Synthetic Aperture Radar (SAR) target recognition using predefined datasets in static scenarios. However, real-world applications demand that models incrementally learn new information without forgetting previously acquired knowledge. The challenge of catastrophic forgetting, where models lose past knowledge when adapting to new tasks, remains a critical issue. In this paper, we introduce IncSAR, an incremental learning framework designed to tackle catastrophic forgetting in SAR target recognition. IncSAR combines the power of a Vision Transformer (ViT) and a custom-designed Convolutional Neural Network (CNN) in a dual-branch architecture, integrated via a late-fusion strategy. Additionally, we explore the use of TinyViT to reduce computational complexity and propose an attention mechanism to dynamically enhance feature representation. To mitigate the speckle noise inherent in SAR images, we employ a denoising module based on a neural network approximation of Robust Principal Component Analysis (RPCA), leveraging a simple neural network for efficient noise reduction in SAR imagery. Moreover, a random projection layer improves the linear separability of features, and a variant of Linear Discriminant Analysis (LDA) decorrelates extracted class prototypes for better generalization. Extensive experiments on the MSTAR, SAR-AIRcraft-1.0, and OpenSARShip benchmark datasets demonstrate that IncSAR significantly outperforms state-of-the-art approaches, achieving a 99.63% average accuracy and a 0.33% performance drop, representing an 89% improvement in retention compared to existing techniques. The source code is available at <uri>https://github.com/geokarant/IncSAR</uri>.https://ieeexplore.ieee.org/document/10838563/Deep learningincremental learningrobust principal component analysis (RPCA)synthetic aperture radar (SAR) target classificationvision transformer |
| spellingShingle | George Karantaidis Athanasios Pantsios Ioannis Kompatsiaris Symeon Papadopoulos IncSAR: A Dual Fusion Incremental Learning Framework for SAR Target Recognition IEEE Access Deep learning incremental learning robust principal component analysis (RPCA) synthetic aperture radar (SAR) target classification vision transformer |
| title | IncSAR: A Dual Fusion Incremental Learning Framework for SAR Target Recognition |
| title_full | IncSAR: A Dual Fusion Incremental Learning Framework for SAR Target Recognition |
| title_fullStr | IncSAR: A Dual Fusion Incremental Learning Framework for SAR Target Recognition |
| title_full_unstemmed | IncSAR: A Dual Fusion Incremental Learning Framework for SAR Target Recognition |
| title_short | IncSAR: A Dual Fusion Incremental Learning Framework for SAR Target Recognition |
| title_sort | incsar a dual fusion incremental learning framework for sar target recognition |
| topic | Deep learning incremental learning robust principal component analysis (RPCA) synthetic aperture radar (SAR) target classification vision transformer |
| url | https://ieeexplore.ieee.org/document/10838563/ |
| work_keys_str_mv | AT georgekarantaidis incsaradualfusionincrementallearningframeworkforsartargetrecognition AT athanasiospantsios incsaradualfusionincrementallearningframeworkforsartargetrecognition AT ioanniskompatsiaris incsaradualfusionincrementallearningframeworkforsartargetrecognition AT symeonpapadopoulos incsaradualfusionincrementallearningframeworkforsartargetrecognition |