Bowen Chen Chen

Dr. Bowen Chen Chen

orcid

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
29 April 2024 in IEEE Transactions on Neural Networks and Learning Systems
Verified
Journal Article
Closed access
XBXueru Bai
Xueru Bai
MYMinjia Yang
Minjia Yang
Bowen Chen ChenBowen Chen Chen
Bowen Chen Chen
FZFeng Zhou
Feng Zhou
Unknown image deformation and few-shot issues have posed significant challenges to inverse synthetic aperture radar (ISAR) target classification. To achieve robust feature representation and precise correlation modeling, this article proposes a novel two-stage few-shot ISAR classification network, dubbed as robust embedding and manifold inference (REMI). In the robust embedding stage, a multihead spatial transformation network (MH-STN) is designed to adjust unknown image deformations from multiple perspectives. Then, the grouped embedding network (GEN) integrates and compresses diverse information by grouped feature extraction, intermediate feature fusion, and global feature embedding. In the manifold inference stage, a masked Gaussian graph attention network (MG-GAT) is devised to capture the irregular manifold of samples in the embedding space. In particular, the node features are described by Gaussian distributions, with interactions guided by the masked attention mechanism. Experimental results on two ISAR datasets demonstrate that REMI significantly improves the performance of few-shot classification and exhibits robustness in various scenarios.

5 August 2024 in IEEE Transactions on Aerospace and Electronic Systems
Verified
Journal Article
Closed access
RXRuihang Xue
Ruihang Xue
XBXueru Bai
Xueru Bai
MYMinjia Yang
Minjia Yang
Bowen Chen ChenBowen Chen Chen
Bowen Chen Chen
FZFeng Zhou
Feng Zhou
Due to the strict observation conditions and special target attributes, inverse synthetic aperture radar (ISAR) may suffer with insufficient number of images for certain space targets, which lead to a considerable decline of the recognition performance. In this article, we propose a robust space target recognition method for sequence ISAR images based on feature distribution transfer learning. To obtain deformation robust sequential features, a sequence homography network (SHN) is firstly proposed and trained by semi-supervised learning. Then the extracted embedding features are aligned and transferred to the class label domain by optimal transport mapping (OTM). Target recognition experiments on few-shot satellite data set illustrate that the proposed method has higher average accuracy and better robustness for scaled, rotated, and combined image deformation.

24 July 2025 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
MYMinjia Yang
Minjia Yang
Bowen Chen ChenBowen Chen Chen
Bowen Chen Chen
YWYue Wang
Yue Wang
XBXueru Bai
Xueru Bai
FZFeng Zhou
Feng Zhou
The collaborated observation of multi-static ISAR can provide comprehensive information for air/space target classification. However, under limited sample conditions, the existing methods is challenged by effective information fusion and distinguishable features extraction. To tackle these issues, a few-shot multi-static ISAR target classification method, dubbed multi-level feature reconstruction (MLRF), is proposed in this article, which comprises two key components: inter-station feature reconstruction (ISFR) and inter-class feature reconstruction (ICFR). Specifically, ISFR facilitates complementary information mining and redundant information suppression by multi-static feature alignment and reconstruction, while ICFR enhances inter-class separability and intra-class compactness via multi-class feature interaction and reconstruction. Additionally, a specialized hybrid loss is designed to ensure desired outputs of the related modules. Experimental results on multi-static ISAR dataset of satellite targets demonstrate that the proposed method significantly improves the few-shot classification accuracy, and ablation studies and visual analysis further validate the effectiveness of each module.

26 January 2026 in Remote Sensing
Verified
Journal Article
Open access
PDF file
Bowen Chen ChenBowen Chen Chen
Bowen Chen Chen
MYMinjia Yang
Minjia Yang
YWYue Wang
Yue Wang
XBXueru Bai
Xueru Bai
Constrained by observation conditions and high inter-class similarity, effective feature extraction and classification of synthetic aperture radar (SAR) targets in few-shot scenarios remains a persistent challenge. To address this issue, this article proposes a few-shot fine-grained SAR target classification method based on multi-scale embedding network and meta-contrastive learning (MSMC). Specifically, the MSMC integrates two complementary training pipelines; the first employs metric-based meta-learning to facilitate few-shot classification, while the second adopts an auxiliary training strategy to enhance feature diversity through contrastive learning. Furthermore, a shared multi-scale embedding network (MSEN) is designed to extract discriminative multi-scale features via adaptive candidate region generation and joint multi-scale embedding. The experimental results on the MSTAR dataset demonstrate that the proposed method achieves superior few-shot fine-grained classification performance compared to existing methods.

4 April 2024 in IET Conference Proceedings
Verified
Journal Article
Closed access
Bowen Chen ChenBowen Chen Chen
Bowen Chen Chen
MYMinjia Yang
Minjia Yang
XBXueru Bai
Xueru Bai
In the field of Synthetic Aperture Radar (SAR), owing to the limitations of observation and imaging conditions, only a limited number of annotated samples are available for model training. Existing deep learning-based methods struggle to effectively extract discernible features, leading to degraded performance or even complete failure. To address these issues, this paper combines prototypical networks from meta-learning with unsupervised contrastive learning. By utilizing meta-learning, the proposed method ensures the model's classification performance under small sample conditions. Additionally, effective SAR data augmentation techniques are designed to incorporate unsupervised contrastive learning, enabling the model to better capture the similarities and differences between classes and enhance its generalization ability. Our proposed approach leverages both the supervised information from class labels and the unsupervised information inherent in the images, allowing the model to deeply explore the high-order information within the data and achieve accurate classification of few shot SAR targets. The recognition results on the MSTAR dataset, consisting of three target classes, demonstrate that the proposed method (CLPN) exhibits higher average recognition accuracy under the small sample condition and robustness to different depression angles.

4 April 2024 in IET Conference Proceedings
Verified
Journal Article
Closed access
MYMinjia Yang
Minjia Yang
Bowen Chen ChenBowen Chen Chen
Bowen Chen Chen
JWJinqi Wang
Jinqi Wang
XBXueru Bai
Xueru Bai
Classification for inverse synthetic aperture radar (ISAR) target remains a challenging task due to insufficient training samples and unknown image deformations. To achieve robust feature extraction and accurate classification, a cross-attention prototype network, namely CAP-Net, is proposed in this article. Different from existing few-shot learning frameworks, the spatial transformation network (STN) is integrated into CAP-Net to achieve adaptive adjustment of unknown deformations, and the cross-attention module (CAM) is also designed to build a bridge between support and query samples for effective information interaction. Experimental results on multi-class ISAR satellite dataset demonstrate that the proposed method can significantly improve the accuracy of few-shot ISAR target classification compared to other state-of-the-art methods.