A5051664121- Advanced Image Processing Techniques
- Image Enhancement Techniques
- Image and Signal Denoising Methods
- Human Pose and Action Recognition
- Advanced Image Fusion Techniques
- Digital Rights Management and Security
- Color Science and Applications
- Video Surveillance and Tracking Methods
- Advanced Graph Neural Networks
- Image and Video Quality Assessment
- Machine Learning in Healthcare
- Remote-Sensing Image Classification
- Digital and Cyber Forensics
- Gait Recognition and Analysis
- Remote Sensing and Land Use
- Advanced Steganography and Watermarking Techniques
- Advanced Neural Network Applications
- Anomaly Detection Techniques and Applications
- Recommender Systems and Techniques
Academy of Broadcasting Science
2023-2024
Communication University of China
2017-2024
Most of the existing low-light enhancement research focuses on global illumination while ignoring issues brightness unevenness and color distortion. To address this dilemma, we propose a image method that can achieve good performance in luminance adjustment correction simultaneously. Specifically, Luminance Adjustment Module is designed to model parameters taking into account relationship between local features, order prevent overexposure or underexposure. Furthermore, design Color...
Detecting small objects is often impeded by blurriness and low resolution, which poses substantial challenges for accurately detecting localizing such objects. Balls players in sports competition videos are more difficult to detect because of motion blur. game that due To tackle this challenge, we propose a object detection backbone network named VHNet based on vertical horizontal information flow scenarios. In order verify the effectiveness our network, manually annotated ball data set...
With the growth of urban population, a series problems have emerged, and how to speed up smart city construction has received extensive attention. Remote sensing images advantages wide spatial coverage rich information, it is suitable for use as research data cities. However, due limitations in imaging sensor conditions complex weather, remote face insufficient resolution cloud occlusion, which cannot meet requirements tasks. The image super-resolution (SR) technique can improve details...
Recently, with the rapid development of mobile sensing technology, capturing scene information by devices in form images or videos has become a prevalent recording method. However, moiré pattern phenomenon may occur when contains digital screens regular strips, which greatly degrade visual performance and image quality. In this paper, considering complexity diversity patterns, we propose novel end-to-end demoiré method, can learn elimination both frequency spatial domains. To be specific,...
In the session-based recommendation algorithm, a better class of methods is to model complex interaction relationship within session as graph structure, and then use neural network capture deep features item from it. However, most models do not deeply mine effective information contained in sequence temporal relationship, nor they pay attention auxiliary contribution global distance different sessions representation. This paper proposes that integrates interactions distance-awareness items...
According to the new format of triple play, video contents interact in integrated broadcast control platform, there are security risks and participation dispute liability trace ability problem, we propose an platform mechanism under play. The builds a bit commitment protocol through hash function, preventing tampering public network transport, using exchanged ElGamal encryption achieve confidentiality comparison event dispute, responsible party. Theoretical analysis shows that this method...
In the past few years, image super-resolution has achieved remarkable progress, but due to texture particularity of remote sensing images, performance in images are not very good. We consider that same object is highly similar, so relevant textures can be transferred from high resolution low image. this paper, we propose a novel network for Specifically, take and reference as inputs, then utilize transformer structure learn transfer which allows model enhance details image, generate...