Yuchen Wang

ORCID: 0000-0002-0408-7508
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Advanced Image Fusion Techniques
  • Remote-Sensing Image Classification
  • Remote Sensing in Agriculture
  • Image Enhancement Techniques
  • Advanced Power Amplifier Design
  • Radio Frequency Integrated Circuit Design
  • Silicon Carbide Semiconductor Technologies
  • Superconducting Materials and Applications
  • Particle accelerators and beam dynamics
  • Microwave Engineering and Waveguides
  • ECG Monitoring and Analysis
  • Network Security and Intrusion Detection
  • GaN-based semiconductor devices and materials
  • Advanced Image Processing Techniques
  • EEG and Brain-Computer Interfaces
  • Advanced Malware Detection Techniques
  • Infrared Target Detection Methodologies
  • Industrial Vision Systems and Defect Detection
  • Internet Traffic Analysis and Secure E-voting
  • Diabetes Management and Education
  • Electrical Fault Detection and Protection
  • Digital Media Forensic Detection
  • Sparse and Compressive Sensing Techniques
  • Heart Rate Variability and Autonomic Control
  • Food Waste Reduction and Sustainability

Yangzhou University
2025

Xi'an Jiaotong University
2024

Xidian University
2019-2024

State Key Laboratory of Electrical Insulation and Power Equipment
2024

University of Science and Technology of China
2024

Institute of Plasma Physics
2024

Macau University of Science and Technology
2024

Zhejiang University of Science and Technology
2022-2023

Beijing University of Chemical Technology
2023

North China Electric Power University
2023

Cloud detection is not only a challenging task, but it also plays major role in image processing. Due to the diversity of cloud and complexity underlying surfaces, most current methods still face great challenges, especially detecting thin cloud. Therefore, we propose method detect pixels GaoFen-1 WFV images. In our method, deep network used learn multi-scale global features so that high-level semantic information obtained process feature learning integrated with low-level spatial order...

10.1016/j.rse.2021.112483 article EN cc-by-nc-nd Remote Sensing of Environment 2021-05-14

Cloud detection is of great significance for the subsequent analysis and application remote-sensing images, it a critical part image preprocessing. In this article, we propose cloud method using convolutional neural networks based on cascaded feature attention channel (CFCA-Net). The CFCA-Net uses module (CFAM) to enhance network toward important color texture feature. CFAM in encoder. CFAN-Net also highlight information dimensions. multi-scale features dilated convolution with different...

10.1109/tgrs.2021.3120752 article EN cc-by IEEE Transactions on Geoscience and Remote Sensing 2021-10-15

In the field of remote sensing image, how to transmit image information more efficiently with limited bandwidth has always been a research hotspot. Compared other ground objects, cloud pixels in are invalid information, so it is meaningful work remove before transmitting and reduce waste useless information. due existence thin clouds complexity underlying surface, most detection algorithms struggle achieve effective separation objects. A deep learning (DL) algorithm based on attention...

10.1109/tgrs.2021.3105424 article EN cc-by IEEE Transactions on Geoscience and Remote Sensing 2021-08-26

The past decades have seen the rapid development of Internet Things (IoT) in various domains. Identifying IoT devices connected to network is a crucial aspect security. However, existing work on identifying based manually extracted features and prior knowledge, leading low efficiency identification accuracy. In this paper, we propose an automatic end-to-end device method (IoT ETEI) CNN+BiLSTM deep learning model, which outperforms traditional methods from perspective overhead identify We...

10.1109/dsc49826.2021.9346251 article EN 2021-01-30

To achieve different compression rates for images while preserving important regions, a semantic network-based deep residual variational auto-encoder is introduced in this paper. The network divided into two components, analysis and an image network. former evaluates the importance of pixels accurately locates regions key information image. According to encoding strategy dynamically adjusted. latter utilizes autoencoder efficiently encode decode images, combining Lagrange multipliers adjust...

10.20944/preprints202501.1786.v1 preprint EN 2025-01-23

Leaf water content (LWC) is a key physiological parameter for assessing maize moisture status, with direct implications crop growth and yield. Accurate LWC estimation essential resource management precision agriculture. This study introduces high-precision method estimating utilizing UAV-based multispectral imagery combined Random Forest Regression (RFR) model. By extracting vegetation indices, image coverage, texture features integrating them ground-truth data, the examines variation in...

10.3390/plants14060973 article EN cc-by Plants 2025-03-20

Abstract Numerical simulation has been widely used to study the temperature field characteristics of muffle sintering, compensating for limitations experiments in obtaining data high-temperature furnaces. We established a numerical model transient high-entropy alloys workpiece furnace and simulated it. The results are agreement with measured results. structure’s influence on workpiece’s uniformity during after sintering process is further discussed. show that increasing height can reduce...

10.1088/1742-6596/3009/1/012076 article EN Journal of Physics Conference Series 2025-05-01

Cloud detection, as a crucial step, has always been hot topic in the field of optical remote sensing image processing. In this paper, we propose deep learning cloud detection Network that is based on Gabor transform and Attention modules with Dark channel subnet (NGAD). This network encoder-decoder framework. The information texture an important feature often used traditional methods. NGAD enhances attention towards features images through proposed extraction module. module larger scale...

10.3390/rs12193261 article EN cc-by Remote Sensing 2020-10-07

Improving diabetes self-management (DSM) is facing real-world challenges among people with type 2 mellitus (T2DM) who have a low education level in resource-limited areas. This study aimed to investigate whether knowledge could predict glycemic levels T2DM rural China. analytical cross-sectional recruited 321 from eight villages by purposive sampling at baseline. After 10 months, 206 patients completed the follow-up survey and HbA1c tests, response rate of 64.17% (206/321). Multiple...

10.1038/s41598-023-45312-y article EN cc-by Scientific Reports 2023-10-25

Pedestrian detection is widely used in cooperative vehicle infrastructure systems. Traditional pedestrian methods perform sufficiently well under sunny scenarios and obtain trustworthy traffic data. However, the drastically decreases rainy scenarios. This study proposes a algorithm with de-raining module that improves accuracy various Specifically, this determines density information of rain effectively removes streaks through module. Then detects pedestrians as pair keypoints to solve...

10.3390/s21010112 article EN cc-by Sensors 2020-12-27

Rapid advances in remote sensing technology have allowed its extensive use defense, land planning, urban traffic monitoring, and natural disaster warning. Remote has penetrated every aspect of modern life. However, some problems need to be solved the data, such as presence clouds images. Efficient airground data transmission can realized by performing cloud rejection on images before satellite transmission. Therefore, this study, were analyzed, an effective detection algorithm was designed....

10.1109/tgrs.2023.3330750 article EN IEEE Transactions on Geoscience and Remote Sensing 2023-01-01

With the development of deep learning, many methods on image denoising have been proposed processing images a fixed scale or multi-scale which is usually implemented by convolution deconvolution. However, excessive scaling may lose detail information, and deeper convolutional network easier to gradient. Diamond Denoising Network (DmDN) in this paper, mainly based meanwhile considering feature information using Diamond-Shaped (DS) module deal with problems above. Experimental results show...

10.1109/vcip49819.2020.9301843 article EN 2022 IEEE International Conference on Visual Communications and Image Processing (VCIP) 2020-12-01

A new-concept Ti-based gas diffusion electrode is successfully developed for hydrogen-related metal electrowinning with low power consumption, high current efficiency, and long service life.

10.1039/d3ta04263d article EN Journal of Materials Chemistry A 2023-01-01

10.1007/s11045-017-0531-7 article EN Multidimensional Systems and Signal Processing 2017-11-20

Currently, smartphones have become an important way for personal privacy leakage. The sensors embedded in low-privileged access policy. Attackers could acquire sensor data without authorization. Therefore, we proposed input recognition attack scheme based on MEMS gyroscope smartphone keyboards. By collecting and analyzing the 3-axis during user clicking screen, a deep learning model is established mapping between data. We call this model. To improve efficiency of scheme, designed window...

10.1109/iccece54139.2022.9712730 article EN 2022 2nd International Conference on Consumer Electronics and Computer Engineering (ICCECE) 2022-01-14

The power supply (PS) of fishtail divertor (FTD) magnet used in experimental advanced superconducting tokamak (EAST) device is multifrequency <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">LC</i> resonant power. In the experiment, change magnetic field will cause current skin effect load divertor, which shift frequency and attenuation current. Therefore, to ensure output amplitude current, feedback control strategy based on identification...

10.1109/tps.2024.3367408 article EN IEEE Transactions on Plasma Science 2024-02-01

For 1500V voltage applications up to hundreds of kilowatts power, integrated power modules adopting various 3-level circuits are widely used. This paper proposes a novel double-sided cooling three level active neutral point clamped (3L-ANPC) SiC MOSFET module with interleaved layout. Benefiting from technology, the commutation loop inductance and junction case thermal resistance (R <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">th-jc</inf> )...

10.1109/apec48139.2024.10509090 article EN 2022 IEEE Applied Power Electronics Conference and Exposition (APEC) 2024-02-25

10.1109/icme57554.2024.10688107 article EN 2022 IEEE International Conference on Multimedia and Expo (ICME) 2024-07-15

Cloud detection is of great significance in the sub-sequent analysis and application remote sensing images, it a critical part image preprocessing. In recent years, convolutional neural network widely used cloud detection. this paper, we introduce Cascaded Feature Attention Network(CFAN), for The CFAN uses cascaded feature attention module(CFAM) to improve extraction capabilities network. Our proposed CFAM contains color module texture module, not only use assist information but also help...

10.23919/icac50006.2021.9594215 article EN 2022 27th International Conference on Automation and Computing (ICAC) 2021-09-02

Trace inspection is a key technology for collecting crime scenes in the criminal investigation department. A lot of information can be obtained by restoring and analyzing remaining traces on scene. However, with development digital technology, trace has become more popular. So, main research this article design realization system based hyperspectral imaging technology. This proposes nondestructive testing Combining basic principles spectroscopy image residual such as car tires, shoe soles,...

10.1155/2022/9524190 article EN Computational Intelligence and Neuroscience 2022-07-15
Coming Soon ...