A5103110660- Blind Source Separation Techniques
- ECG Monitoring and Analysis
- Dental Radiography and Imaging
- Optical measurement and interference techniques
- Industrial Vision Systems and Defect Detection
- EEG and Brain-Computer Interfaces
- Food Supply Chain Traceability
- Surface Roughness and Optical Measurements
- Energy Load and Power Forecasting
- Phonocardiography and Auscultation Techniques
- Piezoelectric Actuators and Control
- Image Processing Techniques and Applications
- Graph Theory and Algorithms
- AI in cancer detection
- Anomaly Detection Techniques and Applications
- Digital and Cyber Forensics
- Iterative Learning Control Systems
- Japanese History and Culture
- COVID-19 diagnosis using AI
- Advanced Measurement and Metrology Techniques
- Spectroscopy and Chemometric Analyses
- Medical Imaging and Analysis
- Cloud Computing and Resource Management
- Image and Object Detection Techniques
- Advanced Neural Network Applications
Guangzhou Academy of Special Equipment Inspection and Testing
2025
Guangdong University of Technology
2017-2024
China Aerospace Science and Technology Corporation
2024
Sheng Jing Hospital
2000
The noninvasive fetal electrocardiogram (FECG) is helpful for well-being monitoring. However, it difficult to obtain high-quality FECG signals because of the maternal (MECG) and noise in abdominal ECG (AECG). To address this problem, an Adaptive Amplitude-Frequency Attention Network (AAFA-Net) proposed extracting from AECG signals, where Frequency Encoder-Decoder (FED) module developed distinguish frequency components Amplitude (AED) devised extract amplitude while Window (WED) designed...
Accurate and rapid diagnosis of COVID-19 using chest X-ray (CXR) plays an important role in large-scale screening epidemic prevention. Unfortunately, identifying from the CXR images is challenging as its radiographic features have a variety complex appearances, such widespread ground-glass opacities diffuse reticular-nodular opacities. To solve this problem, we propose adaptive attention network (AANet), which can adaptively extract characteristic findings infected regions with various...
Electric load forecasting (ELF) is always employed to perform power systems management. However, it difficult predict electric due the following issues: 1) prediction prone external interference, e.g., temperature and weather; 2) user behaviors are random, such as family gatherings business rush orders; 3) consumption varies significantly in different time periods. To solve problems, an adaptive sparse attention network (ASA-Net) proposed for ELF, where spatial (ASSA) module first designed...
<abstract><p>Bone age assessment is of great significance to genetic diagnosis and endocrine diseases. Traditional bone mainly relies on experienced radiologists examine the regions interest in hand radiography, but it time-consuming may even lead a vast error between result reference. The existing computer-aided methods predict based general do not explore specific radiography. This paper aims solve such problems by performing prediction articular surface epiphysis from...
Noninvasive fetal ECG (FECG) is of great significance for monitoring health. However, it challenging to extract FECG signals from the abdominal signal (AECG) due complexity task: 1) are routinely mixed with noise; 2) aliased maternal in time and frequency domain. To solve such problems, an adaptive spectral wavelet network (ASW-Net) proposed extraction, where module, which can improve computational efficiency by replacing convolution operation element-wise Hadamard product domain, first...
Defect detection on magnetic tile surfaces is of great significance for the production monitoring permanent magnet motors. However, it challenging to detect surface defects from due these issues: 1) Defects appear randomly tile; 2) are tiny and often overwhelmed by background. To address such problems, an Adaptive Rotation Attention Network (ARA-Net) proposed defect surface, where Convolution (ARC) module devised capture random learning multi-view feature maps, then Region (RAA) designed...
A critical issue in data replication is to wisely place replicas which involves identifying the best possible nodes duplicate data. Facing dynamics of requests, this paper investigates problem replica placement and update tree networks, where part have pre-existing replicas. We aim develop efficient algorithms accelerate without causing obvious degradation solution quality via reusing Firstly, an heuristic algorithm GRP proposed quickly when users change their requests dynamically, under...
The binary defocusing technique (BDT) shows prospects in high-speed 3D reconstruction. However, the projection mechanism results low accuracy at inappropriate defocus levels due to high-order harmonics, thereby constraining achievable depth range. To overcome problems, this paper proposes a large range binary-focusing technique, aimed fundamentally solving compromise of reconstructed Based on focused-projection strategy, we develop cycle generative adversarial framework for...