XLXingguang Li

Xingguang Li

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Xingguang Li has a total of 16 co-authors. Here are the top co-authors by number of shared publications:

  • Yujian Cai

    School of Electronic Information Engineering, Changchun University of Science and Technology, Changchun 130022, China

    2 shared publications
  • Chen Kang

    School of Electronic Information Engineering, Changchun University of Science and Technology, Changchun, China

    1 shared publication
  • Chen Lei

    School of Electronic Information Engineering, Changchun University of Science and Technology, Changchun, China

    1 shared publication
  • Chunsheng Xu

    Changchun University of Science and Technology, Jilin, Changchun, China

    1 shared publication
  • Guoxiang Liu

    School of Electronic Information Engineering, Changchun University of Science and Technology, Changchun 130022, China

    1 shared publication
  • Guoxiang Liu

    Changchun University of Science and Technology, Jilin, Changchun, China

    1 shared publication

Latest Publications

  • Open Access

    TF-crossnet: a cross-modal attention fusion network for cardiovascular disease classification using pcg and ecg signals

    by Xingguang Li, Yutong Hou, Kaiyao Shi, Yujian Cai

    Electrocardiogram (ECG) and phonocardiogram (PCG) have emerged as crucial non-invasive and portable diagnostic modalities for early cardiovascular disease (CVD) screening. Despite the individual merits of these signal modalities in CVD detection, significant challenges persist, including insufficient inter-modal interaction and suboptimal weight allocation. To address these critical limitations, we proposed a novel Time-Frequency Cross-Modal Attention Fusion Network (TF-CrossNet) designed for precise early CVD diagnosis. The proposed network employs a dual-path multiscale residual structure to extract key time-frequency domain features from PCG and ECG signals, comprehensively capturing multiscale information. Leveraging the intrinsic electro-mechanical coupling relationship of the heart, a bidirectional mutual enhancement attention module is introduced to capture interactive morphological information between PCG and ECG signals, enabling feature-level signal complementation and enhancement. Furthermore, an adaptive fusion strategy based on Bayesian decision theory is developed, establishing a mapping relationship between confidence levels and loss functions to dynamically optimize modal weight allocation. Validated on the 2016 PhysioNet/CinC dataset, the model achieved exceptional performance metrics: 93.13% accuracy, 97.7% specificity, and 98% area under the curve (AUC). Furthermore, comprehensive noise robustness experiments demonstrate that TF-CrossNet maintains superior performance under various noise conditions, achieving an average robustness index of 94.20% compared to existing methods, validating its practical applicability in clinical environments. The superior effectiveness of the proposed approach in CVD classification, providing a novel technological pathway for non-invasive and precision CVD diagnosis.


  • Open Access

    FMCW Radar-Based Human Sitting Posture Detection

    by Guoxiang Liu, Xingguang Li, Chunsheng Xu, Lei Ma, Hongye Li

    Sitting posture is closely related to our health. Poor sitting posture can cause various diseases and harm our physical health. Current methods to detect sitting posture include machine vision, wearable sensors, and pressure sensors. However, these methods have problems with respect to privacy, inconvenience, and cost. In this work, we proposed the use of frequency-modulated continuous wave radar (FMCW) for detecting human sitting posture, which employs wireless signal transmission to enable non-contact detection, protect privacy, and reduce costs. First, the range fast Fourier transform (FFT) and Doppler FFT of the radar’s intermediate frequency (IF) signals are performed to obtain range and Doppler feature information for different sitting postures. Second, to overcome the problem of range FFT bin offset, a single target angle measurement method is proposed to obtain angle features. Subsequently, we constructed various combinations of features to explore the influence of different combinations of features on the detection of posture while sitting. And we used five machine learning algorithms to perform sitting posture detection experiments. Finally, we conducted sedentary experiments in an office setting and provided sitting history records. The experimental results demonstrate that the method we proposed can identify five distinct sitting postures with an average accuracy of 98.07%.


  • Open Access

    Fall Direction Detection in Motion State Based on the FMCW Radar

    by Lei Ma, Xingguang Li, Guoxiang Liu, Yujian Cai

    Accurately detecting falls and providing clear directions for the fall can greatly assist medical staff in promptly developing rescue plans and reducing secondary injuries during transportation to the hospital. In order to facilitate portability and protect people’s privacy, this paper presents a novel method for detecting fall direction during motion using the FMCW radar. We analyze the fall direction in motion based on the correlation between different motion states. The range–time (RT) features and Doppler–time (DT) features of the person from the motion state to the fallen state were obtained by using the FMCW radar. We analyzed the different features of the two states and used a two-branch convolutional neural network (CNN) to detect the falling direction of the person. In order to improve the reliability of the model, this paper presents a pattern feature extraction (PFE) algorithm that effectively eliminates noise and outliers in RT maps and DT maps. The experimental results show that the method proposed in this paper has an identification accuracy of 96.27% for different falling directions, which can accurately identify the falling direction and improve the efficiency of rescue.


  • Top Cited Papers
    Open Access

    Emotion Recognition Using Different Sensors, Emotion Models, Methods and Datasets: A Comprehensive Review

    by Yujian Cai, Xingguang Li, Jinsong Li

    In recent years, the rapid development of sensors and information technology has made it possible for machines to recognize and analyze human emotions. Emotion recognition is an important research direction in various fields. Human emotions have many manifestations. Therefore, emotion recognition can be realized by analyzing facial expressions, speech, behavior, or physiological signals. These signals are collected by different sensors. Correct recognition of human emotions can promote the development of affective computing. Most existing emotion recognition surveys only focus on a single sensor. Therefore, it is more important to compare different sensors or unimodality and multimodality. In this survey, we collect and review more than 200 papers on emotion recognition by literature research methods. We categorize these papers according to different innovations. These articles mainly focus on the methods and datasets used for emotion recognition with different sensors. This survey also provides application examples and developments in emotion recognition. Furthermore, this survey compares the advantages and disadvantages of different sensors for emotion recognition. The proposed survey can help researchers gain a better understanding of existing emotion recognition systems, thus facilitating the selection of suitable sensors, algorithms, and datasets.


  • Open Access

    Radar detection of multi-target vital signs based on blind source separation

    by Jilin Zhang, Xingguang Li, Lingzhe Ma

    Radar non-contact monitoring of vital signs has a broad application prospect in clinical monitoring. Aiming at the problem of strong interference in non-contact vital signs detection (Such as multi-target, random body motion), a blind source separation (BSS) signal detection method based on Fast-ICA is proposed to reduce the interference of multi-target. In this algorithm, entropy is used to evaluate the non Gaussian property, and the appropriate transformation matrix is selected, according to the statistical independence of the signals, the source signals are separated from the observed mixed signals. On this basis, the traditional blind source separation process is improved, and the wavelet transform preprocessing algorithm based on translation invariant is added to suppress the interference of static clutter. The feasibility of this method is verified by simulation experiments.


  • Open Access

    Carrier Frequency Offset Estimation for BPSK Signals in Coherent Optical Communications

    by Chen Kang, Li Xingguang, Zhao Shuang, Chen Lei