A5101813814- Smart Grid Security and Resilience
- Energy Load and Power Forecasting
- Network Security and Intrusion Detection
- Domain Adaptation and Few-Shot Learning
- Image Retrieval and Classification Techniques
- Image Processing and 3D Reconstruction
- IoT and Edge/Fog Computing
- Cloud Computing and Resource Management
- Advanced Data and IoT Technologies
- Advanced Image and Video Retrieval Techniques
- Advanced Algorithms and Applications
- Nonlinear Dynamics and Pattern Formation
- Robotic Path Planning Algorithms
- Advanced Sensor and Control Systems
- Cancer-related molecular mechanisms research
- Speech and Audio Processing
- Online Learning and Analytics
- Advanced Data Compression Techniques
- Human Pose and Action Recognition
- Multimodal Machine Learning Applications
- Coding theory and cryptography
- Distributed Control Multi-Agent Systems
- Advanced Technologies in Various Fields
- Optimization and Search Problems
- Handwritten Text Recognition Techniques
Shantou University
2023-2024
Wuhan University
2023
Wuhan University of Science and Technology
2023
Beijing Microelectronics Technology Institute
2019-2023
Queen's University Belfast
2020-2023
UNSW Sydney
2023
Yunnan Provincial Science and Technology Department
2020-2023
Harbin Normal University
2023
Central South University of Forestry and Technology
2023
Central South University
2023
Vehicular Edge Computing (VEC) is a promising paradigm for autonomous driving. It can reduce delay and energy consumption of tasks. The problem joint task offloading scheduling resource allocation in VEC challenge issue. In this paper, we investigate the offloading, scheduling, VEC, fast changing channel between vehicle an edge server. A target considering time-varying formulated. goal to minimize tasks guarantee Quality Service (QoS) VEC. Constraints on completion time, consumption,...
Abstract Palmprint recognition and palm vein are two emerging biometrics technologies. In the past decades, many traditional methods have been proposed for palmprint recognition, achieved impressive results. However, research on deep learning-based is still very preliminary. this paper, in order to investigate problem of learning based 2D 3D in-depth, we conduct performance evaluation seventeen representative classic convolutional neural networks (CNNs) one database, five databases...
Background: The application of base fertilizer is significant for reducing agricultural costs, non-point source pollution, and increasing crop production. However, the existing fertilization decision methods require many field observations have high prices popularization application. Methods: This study proposes an innovative model integrating machine learning (ML) swarm intelligence search algorithms to overcome above issues. Based on historical data maize, rice, soybean crops, ML including...
The recently proposed learned indexes have attracted much attention as they can adapt to the actual data and query distributions attain better search efficiency. Based on this technique, several existing works build up for multi-dimensional achieve improved performance. A common paradigm of these is (i) map points a one-dimensional space using fixed space-filling curve (SFC) or its variant (ii) then apply indexing techniques. We notice that first step typically uses SFC method, such...
Burst super-resolution has received increased attention in recent years due to its applications mobile photography. By merging information from multiple shifted images of a scene, burst aims recover details which otherwise cannot be obtained using simple input image. This paper reviews the NTIRE 2022 challenge on super-resolution. In challenge, participants were tasked with generating clean RGB image 4× higher resolution, given RAW noisy as input. That is, methods need perform joint...
Traffic characterization (e.g., chat, video) and application identification FTP, Facebook) are two of the more crucial jobs in encrypted network traffic classification. These activities typically carried out separately by existing systems using separate models, significantly adding to difficulty administration. Convolutional Neural Network (CNN) Transformer deep learning-based approaches for CNN is good at extracting local features while ignoring long-distance information from sequence, can...
This study presents a generation expansion planning by incorporating the impacts of renewable energy on mix. The wind-solar power output and its flexibility requirement are integrated into an optimization model to provide realistic representation wind solar resources. is then used for system Jiangsu Province. A comparion demand, electricity price subsidies, carbon emission intensity scenarios reveals scheme path integrating increasing energy. results suggest that installed capacity will...
The Maximum k-Defective Clique Problem (MDCP), as a clique relaxation model, has been used to solve various problems. Because it is hard computational task, previous works can hardly the MDCP for massive sparse graphs derived from real-world applications. In this work, we propose novel branch-and-bound algorithm based on several new techniques. First, two upper bounds of well corresponding reduction rules remove redundant vertices and edges. proposed are particularly useful graphs. Second,...
the instinct characteristics of blockchain technology to write transactions on distributed ledgers offers new opportunities for government improve transparency, prevent fraud, and establish trust in public sector. However, there still exits challenge protect data confidentiality, authenticity ownership when sharing exchanging e-document decentralized network. In order address this concern, we propose a fusion scheme CP-ABE Blockchain called GovChain. GovChain, framework is used implement...
As the information sensing and processing capabilities of IoT devices increase, a large amount data is being generated at edge Industrial (IIoT), which has become strong foundation for distributed Artificial Intelligence (AI) applications. However, most users are reluctant to disclose their due network bandwidth limitations, device energy consumption, privacy requirements. To address this issue, paper introduces an Edge-assisted Federated Learning (EFL) framework, along with incentive...
Photovoltaic power forecasting plays a significant role in the operation of system with high renewables. Owing to powerful data mining ability, artificial intelligence learning based models have achieved impressive success photovoltaic forecasting. However, most these methods cannot be directly employed cases where there is not sufficient historical train reliable model. Fortunately, transfer can utilized address this problem by exploiting knowledge learned from related areas data. In paper,...
With the development of pen-based mobile device, on-line signature verification is gradually becoming a kind important biometrics verification. This thesis proposes method handwritten signatures using both Support Vector Data Description (SVM) and Genetic Algorithm (GA). A 27-parameter feature set including shape dynamic features extracted from data. The genuine each subject are treated as target data to train SVM classifier. As kernel based one-class classifier, can accurately describe...
Power communication network is an important infrastructure of power system. For a large number widely distributed business terminals and terminals. The data protection related to the safe stable operation whole grid. How solve problem that lots nodes need keys avoid situation these cannot exchange information safely because lack keys. In order problem, this paper proposed segmentation combination technology based on quantum key extend limited key. basic idea was obtain division scheme...
Abstract The proportion of distributed photovoltaic grid connection in the distribution network is gradually increasing, and its characteristics high volatility poor stability have brought new challenges to overvoltage problem faced by structure safe stable optimization operation. This paper mainly studies clustering reactive power control Photovoltaic. system architecture cloud edge constructed, voltage autonomy strategy within region after partition proposed. model established with total...
Image rain removal is an important topic in the field of computer vision. In rainy environment, will seriously affect quality imaging, resulting image deformation, blur, poor visibility, and other problems. So outdoor vision system cannot accurately detect object, monitor, works. Therefore, how to effectively eliminate rain-weather interference imaging has a very practical value. absence time series information between frame frame, bottleneck problem technology remove multi-density...
Effective learning hinges on student engagement, yet current assessment methods often rely single-time, subjective self-reports. This study introduced a dynamic, objective approach using facial expression recognition technology to assess comparing it with conventional methods. Participants in synchronous online Chinese as second language classes were analyzed, examining correlations between happiness expressions and six engagement measurements based reports, including self-reported mood,...
Cross-domain learning aims to transfer knowledge learned from one or more datasets other in different domains, so that less data will be required for new tasks and datasets. One big challenge cross-domain is effectively synergize the between domains. In this paper, we propose a solution address using normalizing flow, named as DomainFlow, which works mapping establish sharing source target The flow encourages posterior distributions multi-domain better aligned, leading performance domain...