A5100334644- Advanced Image Processing Techniques
- Image and Signal Denoising Methods
- Advanced Vision and Imaging
- Image Processing Techniques and Applications
- Face and Expression Recognition
- Face recognition and analysis
- Advanced Image and Video Retrieval Techniques
- Advanced Image Fusion Techniques
- Embedded Systems Design Techniques
- Digital Media Forensic Detection
- Advanced Data Storage Technologies
- Video Surveillance and Tracking Methods
- Multimodal Machine Learning Applications
- Distributed and Parallel Computing Systems
- Image Retrieval and Classification Techniques
- Interconnection Networks and Systems
- Advanced Steganography and Watermarking Techniques
- Image Enhancement Techniques
- Parallel Computing and Optimization Techniques
- Generative Adversarial Networks and Image Synthesis
- Anomaly Detection Techniques and Applications
- Biometric Identification and Security
- Chaos-based Image/Signal Encryption
- Distributed systems and fault tolerance
- Network Security and Intrusion Detection
Fujian Normal University
2019-2025
Shenyang Ligong University
2020-2024
Advanced Technology & Materials (China)
2024
Peking University
2024
Hangzhou Normal University
2023-2024
Shandong University
2023
Shandong Academy of Sciences
2023
Qilu University of Technology
2023
Xi'an Polytechnic University
2023
The University of Adelaide
2022-2023
Blind image deblurring aims to recover sharp from a blurred one while the blur kernel is unknown. To solve this ill-posed problem, great amount of priors have been explored and employed in area. In paper, we present blind method based on Local Maximum Gradient (LMG) prior. Our work inspired by simple intuitive observation that maximum value local patch gradient will diminish after process, which proved be true both mathematically empirically. This inherent property process helps us establish...
Recent studies in deepfake detection have yielded promising results when the training and testing face forgeries are from same dataset. However, problem remains challenging one tries to generalize detector created by unseen methods This work addresses generalizable a simple principle: representation should be sensitive diverse types of forgeries. Following this principle, we propose enrich "diversity" synthesizing augmented with pool forgery configurations strengthen "sensitivity" enforcing...
Deep learning is well known as a method to extract hierarchical representations of data. In this paper novel unsupervised deep based methodology, named Local Binary Pattern Network (LBPNet), proposed efficiently and compare high-level over-complete features in multilayer hierarchy. The LBPNet retains the same topology Convolutional Neural (CNN) - one most studied architectures whereas trainable kernels are replaced by off-the-shelf computer vision descriptor (i.e., LBP). This enables achieve...
Recently, deep learning has been successfully applied to the single-image super-resolution (SISR) with remarkable performance. However, most existing methods focus on building a more complex network large number of layers, which can entail heavy computational costs and memory storage. To address this problem, we present lightweight Self-Calibrated Efficient Transformer (SCET) solve problem. The architecture SCET mainly consists self-calibrated module efficient transformer block, where adopts...
Purpose The purpose of this study is to investigate the impact presence a code ethics on quality auditors' judgments, within context new International Standard Quality Controls 1 (ISQC1). Design/methodology/approach A sample 112 professional accountants and auditing students was employed effect (operationalised as vs absence an organisational conduct) audit pertaining inventory writedown, using 2 × full factorial “between‐subjects” experimental design. Findings results indicate that has...
We are dealing with the face cluster recognition problem where there multiple images per subject in both gallery and probe sets. It is never guaranteed to have a clear spatio-temporal relation among of each subject. Considering that image vectors subject, either or probe, span subspace, an algorithm, Dual Linear Regression Classification (DLRC), for developed distance between two subspaces defined as similarity value DLRC attempts find "virtual" located intersection spanning from clusters...
Blind deblurring has received considerable attention in recent years. However, state-of-the-art methods often fail to process saturated blurry images. The main reason is that pixels around regions are not conforming the commonly used linear blur model. Pioneer arts suggest excluding these during process, which sometimes simultaneously removes informative edges and results insufficient information for kernel estimation when large exist. To address this problem, we introduce a new model fit...
Content distribution, especially the distribution of video content, unavoidably consumes bandwidth resource heavily. Internet content providers (ICP) spend lots money to buy network (CDN) service. By deploying thousands edge servers close end users, CDN companies are able distribute efficiently. In lieu traditional systems, we implement a crowdsourcing-based system, Thunder Crystal, which utilizes agents' upload amplify capacity. Agents well motivated contribute storage and system by rebated...
Since deep convolutional neural network (CNN) has achieved excellent results in single image super-resolution (SISR), an increasing number of methods based on CNN have been proposed. Most CNN-based are devoted to finding mapping pixel intensity while ignoring the importance frequency information, which can reflect semantic information images different bands. This leads less effectiveness reconstruction high-frequency details. To address this problem, we propose a novel method named joint...
This study developed machine learning models using different algorithms, including support vector (SVM), random forest (RF), and backpropagation neural network (BPNN), to estimate the critical frequency of F2 layer (foF2) maximum usable for a 3000 km circuit (MUF(3000)F2) based on total electron content (TEC) observed by global navigation satellite system (GNSS) receivers. The ionospheric dataset used comprised TEC, foF2, MUF(3000)F2 measurements from 11 stations in China during solar...
In this paper, we focus on restoring high-resolution facial images under noisy low-resolution scenarios. This problem is a challenging as the most important structures and details of captured are missing. To address problem, propose novel local patch-based face super-resolution (FSR) method via joint learning contextual model. The model based topology consisting sub-patches, which provide more useful structural information than commonly used due to finer patch size. way, models able recover...
In the real world, degradation of images taken under haze can be quite complex, where spatial distribution is varied from image to image. Recent methods adopt deep neural networks recover clean scenes hazy directly. However, due paradox caused by variation captured and fixed parameters current networks, generalization ability recent dehazing on real-world not ideal.To address problem modeling degradation, we propose solve this perceiving density for uneven distribution. We a novel Separable...