A5101825789- Digital Media Forensic Detection
- Advanced Vision and Imaging
- Advanced Steganography and Watermarking Techniques
- Advanced Battery Technologies Research
- Medical Image Segmentation Techniques
- Maritime Transport Emissions and Efficiency
- EEG and Brain-Computer Interfaces
- Advanced Sensor and Control Systems
- Magnetic Field Sensors Techniques
- Image Enhancement Techniques
- Chaos-based Image/Signal Encryption
- Digital Imaging for Blood Diseases
- ECG Monitoring and Analysis
- HVDC Systems and Fault Protection
- Industrial Technology and Control Systems
- COVID-19 diagnosis using AI
- Electrical and Bioimpedance Tomography
- Advanced Algorithms and Applications
- Advanced Image and Video Retrieval Techniques
- Fault Detection and Control Systems
- Hydraulic and Pneumatic Systems
- Face recognition and analysis
- Refrigeration and Air Conditioning Technologies
- Advanced Image Processing Techniques
- Consumer Market Behavior and Pricing
Beijing University of Posts and Telecommunications
2025
Shanghai Maritime University
2006-2024
OriginWater (China)
2024
Gansu Provincial Maternal and Child Health Hospital
2020-2024
Stem Cell Institute
2024
Wellcome/MRC Cambridge Stem Cell Institute
2024
Xi’an Children’s Hospital
2023
Xi'an Jiaotong University
2009-2023
University of Shanghai for Science and Technology
2021-2022
Nanjing University of Aeronautics and Astronautics
2022
Atrial fibrillation (AF) is a serious cardiovascular disease with the phenomenon of beating irregularly. It major cause variety heart diseases, such as myocardial infarction. Automatic AF beat detection still challenging task which needs further exploration. A new framework, combines modified frequency slice wavelet transform (MFSWT) and convolutional neural networks (CNNs), was proposed for automatic identification. MFSWT used to 1 s electrocardiogram (ECG) segments time-frequency images,...
Abstract A hybrid energy system (HES) including hydrogen fuel cell systems (FCS) and a lithium-ion (Li-ion) battery storage (ESS) is established for ships to follow fast load transients. An management strategy (EMS) with hierarchical control presented achieve proper distribution of power enhance stability. In the high-control loop, mechanism based on particle swarm optimization algorithm (PSO) an equivalent consumption minimization (ECMS) proposed. low-level adaptive fuzzy PID controller...
Abstract Tensor convolution is a fundamental operation in convolutional neural networks, especially for processing tensors, which are prevalent real‐world applications. Current methods often convert tensor convolutions into matrix multiplications, leading to data replication, additional memory usage and increased hardware complexity. Here, high‐bit‐efficiency optical accelerator with reduced redundancy lower consumption presented. The bit‐efficiency of the first explored, significantly...
This study is based on breast cancer data from the National Cancer Institute (NCI) database, focusing triple-negative (n = 200) and LumB subtype 400). A generation analysis process combining non-parametric statistics machine learning was designed. In initial stage, wgain algorithm developed by integrating Wasserstein Generative Adversarial Networks (WGAN) Random Forest algorithms. The generated expanded dataset consistent with original data, a Pearson correlation coefficient of approximately...
Abstract Pompe disease (PD) is a rare autosomal recessive disorder that presents with progressive hypertrophic cardiomyopathy. However, the detailed mechanism remains clarified. Herein, PD patient‐specific induced pluripotent stem cells were differentiated into cardiomyocytes (PD‐iCMs) exhibited cardiomyopathic features of PD, including decreased acid alpha‐glucosidase activity, lysosomal glycogen accumulation and hypertrophy. The defective mitochondria involved in cardiac pathology as shown...
Smart agriculture represents a significant trend in agricultural development, given its potential to enhance operational efficiency and reduce labor intensity. Despite the adoption of modern greenhouse technologies, such as sensors automation systems, crop transportation is still largely achieved through manual labor, due complex environment narrow terrain greenhouses. To address this challenge, work proposes design an intelligent vehicle that capable transporting crops commercial...
Abstract The fuel cell system (FCS) is commonly combined with an energy storage (ESS) for enhancing the performance of ship. Consequently, battery ESS size and power allocation strategy are critical hybrid system. This paper focuses on designing a method to solve these two problems. First, degradation model employed assess lifetime. Subsequently, sizing problem optimal integrated into cost‐minimization problem, which solved by double‐loop optimization approach. inside loop utilizes calculate...
This paper presents an identification algorithm of the scalar Preisach model by using a set symmetric minor hysteresis loops and suggests improved vector to describe anisotropic behavior non-oriented (NO) electrical steel sheet (ESS) under rotating magnetic fields. Symmetric loop is quite easy measured, so based on instead first-order reverse curve presented in this paper. The employs nonlinear coefficient, which function magnitude direction flux density, property NO ESS increases numerical...
In this paper, multiple autonomous underwater vehicle (multi-AUV) formation control with obstacle avoidance ability in 3D complex environments based on an event-triggered model predictive (EMPC) is proposed. Firstly, multi-AUV motion systems are developed. The navigation reference trajectory of the follower AUVs can be obtained using a relative model. Secondly, order to overcome speed jump and problem systems, compatibility constraints presented MPC that limit uncertainty deviation each AUV....
The steering mechanism of ship gear is generally driven by a hydraulic system. precise control the cylinder in can be achieved target rudder angle. However, systems are often described as nonlinear with uncertainties. Since system parameters uncertain and performances influenced disturbances noises, robustness cannot satisfied approximating theory linear theory. In this paper, learning-based model predictive controller (LB-MPC) designed for position an electro-hydraulic order to reduce...
Background: Placental mesenchymal dysplasia (PMD) is a rare placental vascular malformation of unknown etiology. PMD may coexist with healthy fetus, and its ultrasound appearance similar to that hydatidiform mole, especially the partial type. Prenatal ultrasonography vital for accurate diagnosis these conditions. This study aimed summarize characteristics prenatal ultrasonographic images across different gestational weeks (W) evaluate analyze factors influence pregnancy outcomes related PMD....
Powell's direction set method (PDSM) is widely used in image registration. However, its performance found to be strongly dependent on the initial solution. In this paper we propose a novel combination of PDSM with initialization by differential evolution genetic algorithm. To further improve convergence, multi-resolution approach taken. Mutual information chosen as error metric for investigation. The results presented show that proposed does significantly and it efficient robust.
The use of green energy to power ships in the marine industry has attracted increasing attention recent years. This paper presents an inland river cruise ship supplied by a fuel cell (FC) as main source and supercapacitor (SC) auxiliary source. Its propulsion inverter adopts proposed high-boost Z-source inverter, high-voltage-boost (HVB-ZSI) principle is studied. advantages this HVB-ZSI two cases are verified through simulation. In case 1, it can be seen that capacitance voltage only 250 V,...
Most of the face anti-spoofing methods improve generalization capability by adversarial domain adaptation via training source and target data jointly. However, considering privacy, it is impractical in application. Hence, we propose a data-free adaptative framework to optimize network without using labeled modeling into problem learning with noisy labels. To obtain more reliable pseudo labels, dynamic images background capture motion divergences between real attack faces. Nonetheless,...
Wide bandwidth is usually required to transmit depth-image-based three-dimensional (3-D) video. An effective way reduce the requirement converting progressive 3-D video an interlaced one. Since cannot be displayed on end directly, must deinterlaced. A deinterlacing algorithm for proposed. The first performs reference image and its depth map by intrafield interpolation based direction. works well in eliminating jagged effect adaptively selecting direction of interpolation; it also provides...