Lingfeng Chen

I am admitted to the College of Electronic Science and Technology, National University of Defense Technology (NUDT), Changsha, China to persue my Ph.D degree in Information and Communication Technology, supervised by Prof. Liu and Dr. Hu. I majored in Microelectronic Science and Engineering in my undergraduate years. As a scientific researcher, I am extremely interested in cross-displine topics and atificial intelligence.

During my undergrad years, I was also interested in Syntheic Biology and I had participated 2 years in International Genetically Engineered Machine competition (iGEM) helded originally by Massachusetts Institute of Technology (MIT), as the student leader of team NUDT_CHINA 2022, NUDT_CHINA 2023. We had won the Gold Medal of the contest for 2 years as well as nomination of the Best Therapeutic Project in 2023.

My research interest includes Radar Automatic Target Recogntion (RATR), Deep Learning (DL), meta-learning, Graph Neural Network (GNN), eXplainable AI (XAI).

Email: chenlingfeng@nudt.edu.cn  /  ORCID: 0009-0003-2690-9407
/ GitHub: MountainChenCad

Mar 2025: Happy to announce that our paper Human Motion Recognition through Multi-Aspect mmWave FMCW Radar Data got accepted to IGARSS 2025!

Jan 2025: My National College Student Innovation and Entrepreneurship Project, "A Deep Graph Model of Virus-Human Protein-to-Protein Interaction". Rated as Excellent (rank 1/38 in 2024 projects).

Dec 2024: The codes of Channel-DN4 are now released!

Nov 2024: We are in the final of the Challenge Cup! A contest for multiModal detection through IR&RD data. Awarded with Honorable Mention (rank 8/61 in our track).

Nov 2024: Happy to announce that our paper HRRPGraphNet: Make HRRPs to Be Graphs for Efficient Target Recognition got accepted to IET Electronics Letters!

Oct 2024: The codes of GAF-MLGNN are now released!

Sep 2024: The codes of HRRPGraphNet are now released!

GAF-MLGNN: An Efficient Meta-Learning Framework for Few-shot HRRP RATR with GNN
Lingfeng Chen, Panhe Hu*, Qi Liu, Zhen Liu
Submitted to IEEE Transactions on Signal and Information Processing over Networks (TSIPN), 2025

To efficiently utilize the target information from HRRP samples under few-shot scenario, our solution brings a new mindset to take both inner-sample and inter-sample information into consideration. More importantly, we have proposed a novel task set-based meta learning method for GNN, which further enhances the generalization ability of the model.

Few-shot Human Motion Recognition through Multi-Aspect mmWave FMCW Radar Data
Hao Fan, Lingfeng Chen*, Chengbai Xu, Jiadong Zhou, Yongpeng Dai, Panhe Hu
IEEE International Symposium on Geoscience and Remote Sensing (IGARSS), 2025
Arxiv

Radar human motion recognition methods based on deep learning models has been a heated spot of remote sensing in recent years, yet the existing methods are mostly radial-oriented. In practical application, the test data could be multi-aspect and the sample number of each motion could be very limited, causing model overfitting and reduced recognition accuracy. This paper proposed channel-DN4, a multi-aspect few-shot human motion recognition method... First, local descriptors are introduced for a precise classification metric. Moreover, episodic training strategy was adopted to reduce model overfitting. To utilize the invariant semantic information in multi-aspect conditions, we considered channel attention after the embedding network to obtain precise implicit high-dimensional representation of semantic information. We tested the performance of channel-DN4 and methods for comparison on measured mmWave FMCW radar data. Read More

HRRPGraphNet: Make HRRPs to Be Graphs for Efficient Target Recognition
Lingfeng Chen, Xiao Sun, Zhiliang Pan, Zehao Wang, Xiaolong Su, Zhen Liu, Panhe Hu*
IET Electronics Letters (ELL), 2024
Arxiv / Authorea / Wiley Link

High Resolution Range Profiles (HRRPs) have become a key area of focus in the domain of Radar Automatic Target Recognition (RATR). Despite the success of deep learning based HRRP recognition, these methods need a large amount of training samples to generate good performance, which could be a severe challenge under non-cooperative circumstances... Currently, deep learning-based models treat HRRPs as sequences, which may lead to ignorance of the internal relationship of range cells. This letter proposes HRRPGraphNet, a novel graph-theoretic approach, whose primary innovation is the use of the graph-theory of HRRP which models the spatial relationships among range cells through a range cell amplitude-based node vector and a range-relative adjacency matrix, enabling efficient extraction of both local and global features in noneuclidean space. Experiments on the aircraft electromagnetic simulation dataset confirmed HRRPGraphNet's superior accuracy and robustness compared with existing methods, particularly in limited training sample condition. This underscores the potential of graph-driven innovations in enhancing HRRP-based RATR, offering a significant advancement over sequence-based methods. Read More

A Deep Learning-Based Target Radial Length Estimation Method through HRRP Sequence
Lingfeng Chen, Panhe Hu*, Zhiliang Pan, Xiao Sun, Zehao Wang
IEEE Asia-Pacific Conference on Antennas and Propagation (APCAP), 2024
Arxiv / IEEE Explore

This paper introduces an innovative deep learning-based method for end-to-end target radial length estimation from HRRP (High Resolution Range Profile) sequences. Firstly, the HRRP sequences are normalized and transformed into GAF (Gram Angular Field) images to effectively capture and utilize the temporal information... Subsequently, these GAF images serve as the input for a pretrained ResNet-101 model, which is then fine-tuned for target radial length estimation. The simulation results show that compared to traditional threshold methods and simple networks such as one-dimensional CNN (Convolutional Neural Network), the proposed method demonstrates superior noise resistance and higher accuracy under low SNR (Signal-to-Noise Ratio) conditions. Read More

Thank Dr. Jon Barron for sharing the source code of the website.