A5037630663- Wireless Networks and Protocols
- Vehicle License Plate Recognition
- Advanced Neural Network Applications
- Handwritten Text Recognition Techniques
- Cooperative Communication and Network Coding
- Network Traffic and Congestion Control
- Advanced Biosensing Techniques and Applications
- Network Security and Intrusion Detection
- Complex Systems and Time Series Analysis
- Gene expression and cancer classification
- Video Surveillance and Tracking Methods
- Energy Load and Power Forecasting
- Mechanical and Optical Resonators
- Advanced Image and Video Retrieval Techniques
- Stock Market Forecasting Methods
- Access Control and Trust
- Security and Verification in Computing
- Cloud Data Security Solutions
- Generative Adversarial Networks and Image Synthesis
- Advanced Chemical Sensor Technologies
- Advanced Image Processing Techniques
- Advanced MEMS and NEMS Technologies
- Cloud Computing and Resource Management
- Software-Defined Networks and 5G
- Face recognition and analysis
Beihang University
2020-2024
Université Libre de Bruxelles
2019
Beijing University of Technology
2018-2019
Peking University
2014
Cloud Computing Center
2012
Automatic license plate recognition plays an important role in intelligent transportation systems and is of great significance. However, at present, most current approaches are only concerned under restrictive conditions, where the plates shot a frontal view good light conditions. These not robust enough real-world complex capture scenarios, such as uneven condition or oblique shooting angle. In order to improve robustness recognizing detection network (CA-CenterNet) proposed this paper,...
Cloud computing is very useful for big data owner who doesn't want to manage IT infrastructure and technique details. However, it hard trust multi-layer outsourced system in cloud environment verify which service leads the problem. Similarly, provider cannot simply computation applications. At last, verification itself may also leak sensitive information from owner. We propose a new three-level definition of verification, threat model, corresponding trusted policies based on different roles...
There are a large number of studies on geospatial object detection. However, many existing methods only focus either accuracy or speed. Methods with both fast speed and high great importance in some scenes, like search rescue, military information acquisition. In remote sensing images, there targets that small have few textures low contrast compared the background, which impose challenges this paper, we propose an accurate single shot detector (AF-SSD) for spatial imagery to solve these...
Quantitative investment which combines financial data with mathematics and computer technology has become a new rising method of the international community in recent ten years. This paper proposes quantitative prediction model based on latest mathematical theory analysis method. Firstly, we construct whole system. Then do qualitative market multi-fractal to see whether there exist fractal characteristics. Finally, use support vector machine (SVM) predict changes assets. We choose Shanghai...
A novel framework is proposed that combines multi-resonance biosensors with machine learning (ML) to significantly enhance the accuracy of parameter prediction in biosensing. Unlike traditional single-resonance systems, which are limited one-dimensional datasets, this approach leverages multi-dimensional data generated by a custom-designed nanostructure, periodic array silicon nanorods triangular cross-section over an aluminum reflector. High bulk sensitivity values achieved for...
This paper proposes a new method of TCP with network coding (NC) and implements the version (TCP-NC) on Linux kernel. In TCP-NC design, it introduces layer, constructs encoding, header format, encoding queue, ACK-NC, redundancy coefficient timer. kernel implementation, does some simplification: setting length queue (equals two packets), making ACK-NC instead ACK-TCP so on. Then, this makes changes to core functions data structures, including modifying transmission control block, changing...
We propose a method of dynamically adjusting redundancy parameter, which can be applied in "Network Coding Meets TCP". It help solve the packet losses TCP layer and part new protocol future Internet. In our scheme, we mainly introduce "time-cache" "time-classification management algorithm" to network coding sender. By analyzing time difference between acknowledgements given by sink, give right value parameter using time-classification algorithm, for choosing is very important mask from TCP....
Face swapping in videos usually has strong entertainment applications. Deep Fakes (in Faces) are a recent topic deep learning where the main idea is to substitute face of person video with another person. But one drawbacks method that between two successive frames there inconsistencies faces, such as changing color, flickering or eyebrows change. In this paper, we propose convolutional neural network for faces based on autoencoders which share same encoder. network, encoder can distinguish...