A5115044980- Image Enhancement Techniques
- Visual Attention and Saliency Detection
- Blockchain Technology Applications and Security
- Advanced Image Fusion Techniques
- Remote Sensing and Land Use
- Microgrid Control and Optimization
- Brain Tumor Detection and Classification
- Control Systems and Identification
- Advanced Decision-Making Techniques
- Iterative Learning Control Systems
- Energy, Environment, Economic Growth
- Remote-Sensing Image Classification
- EEG and Brain-Computer Interfaces
- Cognitive Science and Education Research
- Energy, Environment, and Transportation Policies
- Machine Learning in Bioinformatics
- Infrared Target Detection Methodologies
- Environmental Impact and Sustainability
- Frequency Control in Power Systems
- Advanced Computational Techniques and Applications
- Advanced Control Systems Optimization
- HVDC Systems and Fault Protection
City University of Hong Kong, Shenzhen Research Institute
2024
Central South University
2016-2024
State Grid Corporation of China (China)
2022
Beijing Academy of Artificial Intelligence
2020
Camouflaged object detection has been considered a challenging task due to its inherent similarity and interference from background noise. It requires accurate identification of targets that blend seamlessly with the environment at pixel level. Although existing methods have achieved considerable success, they still face two key problems. The first one is difficulty in removing texture noise thus obtaining edge frequency domain information, leading poor performance when dealing complex...
In order to analyze the factors that affect energy consumption in Hunan Province comprehensively, firstly, this paper used Logarithmic Mean Divisia Index (LMDI) and divided total growth of three industries from 2006 2015 into scale effect, structure effect efficiency effect. Secondly, long-term alternatives planning (LEAP) system was used, LEAP-Hunan model with benchmark, comprehensive adjustment scenarios set up impacts kinds effects on profoundly. LMDI qualitative decomposition results...
Detecting camouflaged objects is expected to be a challenging task due the hard-distinguihsed boundaries of targets. Although existing learning-based methods have concentrated on utilizing boundary information enhance object detection, absence difficulty estimation causes them treat all regions as equal, thereby making it more distinguish high intrinsic similarity regions. To address this issue, by filtering redundant easy boundaries, we proposed Edge Attention Network (EANet) extract...
This article first introduced the process of renewable energy power trading, sorted out core information and data required for traceability entire chain from generation side to consumption side, established mechanism "sub-link + overall calculation verification". Using blockchain encryption anti-tampering features, a blockchain-based collection system was studied designed, which realized on-chain certification verification key ensure authenticity credibility traceable data, improve...
With the rapid increase of photovoltaic (PV) penetration and distributed grid access, generation (PVG)-integrated multi-area power systems may be disturbed by more uncertain factors, such as PVG, grid-tie inverter parameters, resonance. These factors will exacerbate frequency fluctuations PVG integrated interconnected systems. For system, this paper proposes a load control (LFC) strategy based on double equivalent-input-disturbance (EID) controllers. The linear model system were established,...
This paper is intended to solve the problem of stochastic disturbance suppression for heating furnace temperature control existed in billet process. It aims propose an active compensation method based on adaptive reset state observer (ARSO) suppression. The structure system (ACCS) presented and ARSO designed. In this method, equivalent input (EID) drawn into channel ACCS eliminate influence by real-time estimation outside disturbance. With consider strong nonlinear, large time delay, big...
Aiming at the problem that traditional one - dimensional linear processing method can't effectively extract characteristic information of weak signal brain, this paper presents a lesion location algorithm electroencephalogram (EEG) based on image features. Image features is introduced into epileptic processing, which can feature and locate lesions in EEG signals. First all, theoretical basis onedimensional reconstruction two-dimensional described paper. Then, order to make have...