Article
Environmental Sciences
Ting Chen, Mengni Liu, Tao Gao, Peng Cheng, Shaohui Mei, Yonghui Li
Summary: In this study, a novel fusion-based defogging algorithm is proposed to address the limitations of traditional dark channel method and improve the restoration quality of defogged images in large sky areas. Experimental results demonstrate the superior performance of the proposed algorithm on UAV foggy images.
Article
Geochemistry & Geophysics
Xudong Kang, Zhengyao Fei, Puhong Duan, Shutao Li
Summary: This study developed a novel fog model to remove fog from hyperspectral images, achieving high-quality defogging by calculating fog density map and estimating haze abundance in different spectral bands. Experimental results demonstrate that the proposed method outperforms other approaches in computer vision and remote sensing fields in terms of dehazing performance.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Chemistry, Analytical
Feng Ling, Yan Zhang, Zhiguang Shi, Jinghua Zhang, Yu Zhang, Yi Zhang
Summary: This paper introduces a polarized image defogging algorithm based on sky segmentation results and transmission map optimization. The algorithm effectively removes the halo effect in the reconstructed image of large area sky region through joint sky segmentation and three-step transmission optimization method.
Article
Chemistry, Analytical
Sheng-Wei Cheng, Yi-Ting Lin, Yan-Tsung Peng
Summary: Bilateral Filtering (BF) is an effective technique for edge-preserving smoothing in image processing, but it struggles with differentiating noise and details in image denoising. This letter proposes a novel Dual-Histogram BF (DHBF) method that utilizes a noise-reduced guidance image to compute the range kernel, removing noisy pixels for improved denoising results. Experimental results show that DHBF outperforms other state-of-the-art BF methods.
Article
Geochemistry & Geophysics
Bo-Hao Chen, Hsiang-Yin Cheng, Yi-Syuan Tseng, Jia-Li Yin
Summary: In this letter, a two-pass (TP) BF and an adaptive control scheme of range kernels for noise-invariant edge-preserving image smoothing are proposed. Experimental results show that the TP-based BF outperforms existing bilateral filters in terms of both feature- and gradient-aware measures.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2022)
Article
Computer Science, Software Engineering
Pu-Cheng Zhou, Ying Xue, Mo-Gen Xue
Summary: This study presents a novel edge-preserving image smoothing algorithm using adaptive side window joint bilateral filtering. By optimizing the position and size of the filtering kernel, it effectively addresses issues such as halo artifact and texture removal.
Article
Computer Science, Software Engineering
Lixi Jiang, Xujie Li, Yandan Wang
Summary: Texture filtering technique aims to remove insignificant textures and retain important structures. In this paper, we propose an effective unsupervised deep bilateral texture filtering neural network that achieves texture smoothing. The model is trained with a bilateral texture loss function and does not require ground truth smoothing images. Extensive experiments show that our approach outperforms existing methods in effectively removing textures while preserving the main image structures.
Article
Chemistry, Multidisciplinary
Lisang Liu, Chengyang Ke, He Lin
Summary: This paper proposes a novel Dark-Center algorithm, which is a joint learning framework based on image defogging and target detection, aiming to solve the problems of low visibility and unclear targets in foggy environments.
APPLIED SCIENCES-BASEL
(2023)
Article
Instruments & Instrumentation
Guote Liu, Jinhui Zhou, Tong Li, Weiquan Wu, Fang Guo, Bing Luo, Sijun Chen
Summary: The algorithm decomposes input source images using hybrid curvature filtering and adopts different fusion strategies based on the attributes of the image layers, resulting in a fusion image that is more conducive to human visual perception and computer analysis compared to other algorithms.
INFRARED PHYSICS & TECHNOLOGY
(2022)
Article
Computer Science, Artificial Intelligence
Liang Dong, Ding Gangyi, Yan Dapeng, Huang Kexiang
Summary: Under the leadership, contemporary digital media have been promoted, utilizing artificial intelligence technology for video image capture, processing, and classification. Image processing technology and digital media technology play a crucial role in the application during the current era of short video outbreak. However, many short videos suffer from low definition, leading to unclear captured images, which adversely affect image recognition and resolution. The combination of image dehazing algorithm and artificial intelligence enhances image processing in digital media and facilitates the protection and classification of image data.
Article
Computer Science, Software Engineering
Hai Yao, Huawang Qin, Qian Wu, Zhisong Bi, Xuezhu Wang
Summary: In this paper, a new defogging technique for multi-exposure images combined with prior algorithm is proposed. The algorithm calculates the transmittance of different regions of the haze image and decomposes and fuses the image using guided filtering. The algorithm has better performance in haze removal.
Article
Computer Science, Information Systems
Qiwei Xing, Chunyi Chen, Zhihua Li
Summary: This study introduces a novel denoising algorithm framework that improves image quality through adaptive sampling and an improved bilateral filtering algorithm.
MULTIMEDIA TOOLS AND APPLICATIONS
(2021)
Article
Chemistry, Analytical
Vadim Ziyadinov, Maxim Tereshonok
Summary: This paper continues the research on the properties of convolutional neural network-based image recognition systems and explores ways to enhance the systems' immunity to noise and robustness. It focuses on adversarial attacks, which are not easily perceptible to human eyes but greatly reduce the accuracy of the neural networks. The paper proposes a technique that utilizes low-pass image filtering to mitigate the influence of high-frequency distortions caused by adversarial attacks, thereby improving image recognition accuracy. The technique is resource-efficient and easy to implement, bridging the gap between artificial neural networks and human recognition logic.
Article
Engineering, Electrical & Electronic
Lei He, Yongfang Xie, Shiwen Xie, Zhipeng Chen
Summary: This article proposes a structure-preserving texture smoothing method based on scale-aware bilateral total variation. The method utilizes a joint bilateral filter to construct the window bilateral variation and combines it with the window total variation to form a regularizer called bilateral total variation. This regularizer accurately quantifies the characteristics of texture and structure, allowing for the fine smoothing of salient textures while preserving weak edge structures and small structures. Experimental results demonstrate the superiority of the proposed method in texture smoothing and other applications.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2023)
Article
Automation & Control Systems
Edgar Estrada, Wen Yu, Xiaoou Li
Summary: In this study, the stability of bilateral teleoperation with haptic feedback under phase transition between constraint and unconstraint motions is proven using Lyapunov-Krasovskii method and hybrid systems theory. The conditions of the theoretical results are more general, including time-varying delays, nonlinear systems, force feedback, and phase transition. The application results demonstrate the efficiency of the proposed control method using admittance control.
JOURNAL OF THE FRANKLIN INSTITUTE-ENGINEERING AND APPLIED MATHEMATICS
(2021)
Article
Computer Science, Information Systems
Wei Liang, Songyou Xie, Dafang Zhang, Xiong Li, Kuan-ching Li
Summary: The Industrial Internet of Things (IIoT) aims to improve economic benefits by optimizing process controls, but the use of Radio Frequency Identification (RFID) technology in IIoT raises concerns about security and cost. To address these issues, a mutual authentication scheme incorporating Deep Learning techniques is proposed for secure access authentication of IC circuits on the IoT.
ACM TRANSACTIONS ON INTERNET TECHNOLOGY
(2022)
Article
Computer Science, Hardware & Architecture
Dun Li, Dezhi Han, Noel Crespi, Roberto Minerva, Kuan-Ching Li
Summary: Supply chain finance provides credit for small and medium-sized enterprises with low credit lines and small financing scales. This study proposes a Blockchain-based secure storage system, Fabric-SCF, to achieve data security and fine-grained access control. Experimental results show that Fabric-SCF performs efficiently in a simulated real-world operating scenario.
JOURNAL OF SUPERCOMPUTING
(2023)
Article
Computer Science, Information Systems
Fang Feng, Kuan-Ching Li, Erfu Yang, Qingguo Zhou, Lihong Han, Amir Hussain, Mingjiang Cai
Summary: This article presents a novel hybrid algorithm NBG, which improves the performance of imbalanced classifications by combining oversampling and a feature selection algorithm. The proposed NBG algorithm significantly outperforms existing and recently published algorithms in the classification of imbalanced small-sample data sets.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Yifan Chen, Shaomiao Chen, Kuan-Ching Li, Wei Liang, Zhiyong Li
Summary: Mobile edge computing (MEC) enhances computation capabilities of smart mobile devices for computation-intensive applications via efficient computation offloading. However, limitations in wireless and computational resources often hinder MEC development. To address offloading in time-varying wireless networks, offloading decisions and resource allocation must be jointly managed. Traditional optimization methods struggle with combinatorial optimization in dynamic network environments. Therefore, we propose a deep reinforcement learning (DRL)-based approach, named DRJOA, for optimizing offloading decisions, computation, and wireless resource allocation. Simulation results demonstrate that DRJOA outperforms benchmark methods in offloading decisions and system utility.
CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS
(2023)
Retraction
Computer Science, Information Systems
Qian Zhang, Maozhu Jin, Yanan Wang, Xia Lei
CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS
(2022)
Article
Economics
Dun Li, Dezhi Han, Zibin Zheng, Tien-Hsiung Weng, Kuan-Ching Li, Ming Li, Shaokang Cai
Summary: Short and Distort (S & D) is a prevalent price manipulation scheme in the futures trading market, especially in the context of Bitcoin futures. This article presents the first detailed empirical study on S & D, combining the analysis of available information and cryptocurrencies, proposing a problematic definition and discriminatory criteria for S & D in Bitcoin futures. Additionally, a model for real-time detection of S & D in perpetual and term contracts is developed, demonstrating higher accuracy and robustness compared to previous studies. The results highlight significant S & D manipulation in both perpetual and term contracts, with Binance exchange being relatively secure and Bittrex exchange being the most vulnerable to S & D.
COMPUTATIONAL ECONOMICS
(2023)
Article
Engineering, Electrical & Electronic
Lijun Xiao, Dezhi Han, Tien-Hsiung Weng, Shaomiao Chen, Han Deng, Alireza Souri, Kuan-Ching Li
Summary: Digital twin network (DTN) is crucial for efficient optimization in modern networks as it has real-time data and interacts with the physical network. This research proposes an evolutionary framework for server layout optimization, which improves the efficiency of evolutionary algorithms and reduces computational cost. Offline-learning and self-examining guided local search methods are used to reduce search space and improve search efficiency. A look-up table-based hybrid approach is used for solution evaluation, reducing computational overhead. Experimental results demonstrate significant improvements in search efficiency and convergence performance.
INTERNATIONAL JOURNAL OF COMMUNICATION SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Lijun Xiao, Dezhi Han, Dun Li, Wei Liang, Ce Yang, Kuan-Ching Li, Arcangelo Castiglione
Summary: With the advancements in computing and networking technologies, there has been a rise in cryptocurrencies or digital tokens, offering various payment services with different costs, quality, and safety. Analyzing blockchain transaction data and historical transaction trends can help identify illegal behaviors, such as money laundering, at an early stage. In this article, a novel method called CTDM is proposed, which combines EvolveGCN with MGU and global representations to achieve better performance in abnormal transaction detection compared to existing methods.
Article
Computer Science, Information Systems
Teng Ma, Yue Xiao, Xia Lei
Summary: This letter discusses a novel one-snapshot multiple signal classification (MUSIC)-derived algorithm based on channel reconstruction for joint angle-of-arrival (AoA) and angle-of-departure (AoD) estimation in multi-input multi-output (MIMO) arrays. Compared to conventional two-dimensional MUSIC (2D-MUSIC), the proposed algorithm not only avoids the problem of matrix rank deficit but also achieves significant complexity reduction while maintaining satisfactory accuracy. Simulation results confirm the utility and advancement of the algorithm, particularly in scenarios with high real-time requirements.
IEEE WIRELESS COMMUNICATIONS LETTERS
(2023)
Review
Mathematics
Sisi Zhou, Kuanching Li, Lijun Xiao, Jiahong Cai, Wei Liang, Arcangelo Castiglione
Summary: Since the birth of Bitcoin, blockchain has evolved from a critical cryptocurrency technology to an enabling technology, revolutionizing fields requiring credibility and high-quality data for decision making. In business intelligence and business process management, users can build their blockchain-enabled collaboration and data-sharing ecosystem. This paper presents the development process of blockchain and introduces important consensus mechanisms, providing a comparison and analysis of these mechanisms to lay a foundation for selecting appropriate consensus mechanisms for different scenarios and application fields.
Article
Computer Science, Interdisciplinary Applications
Zhijie Sun, Dezhi Han, Dun Li, Tien-Hsiung Weng, Kuan-Ching Li, Xiaojun Mei
Summary: This article proposes a blockchain-based solution for medical record data storage and sharing, utilizing Hyperledger fabric blockchain and encryption technology to ensure the security and shareability of medical records. Experiments have shown that this solution outperforms existing approaches.
COMPUTERS & INDUSTRIAL ENGINEERING
(2023)
Article
Physics, Multidisciplinary
Pei Li, Zhiming Lin, Kuanching Li, Subhash Bhalla
Summary: In this study, a message spreading model for social networks is proposed, which introduces decaying attention and modifies the duplicate forwarding model. User influence is defined and theoretically analyzed through generating functions based on this model. The diffusion threshold is further investigated and its existence and uniqueness are proved. Simulation results demonstrate the efficiency of user influence in this model, and analysis of user influence and diffusion thresholds show its accuracy and promising potential.
PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS
(2023)
Article
Computer Science, Hardware & Architecture
Dun Li, Dezhi Han, Tien-Hsiung Weng, Zibin Zheng, Hongzhi Li, Kuan-Ching Li
Summary: Stablecoins have facilitated the growth of decentralized payments and the emergence of a new generation of payment systems using cryptocurrencies and Blockchain technology. However, the existing research lacks a comprehensive overview of Stablecoins that focuses on their full context, stabilization mechanisms, and payment applicability. This paper provides a thorough summary of the definition, current state, and ecosystem of Stablecoins. It discusses the system structure, stability mechanisms, and their applicability in payment scenarios. The study identifies asset-backed Stablecoins as the most efficient and widely used, while cryptocurrency-backed Stablecoins are more balanced in relation to the original concept. Algorithm-backed Stablecoins show significant potential for development but are hesitant due to the lack of collateral or deposit reserves, making them prone to collapse. The paper concludes by presenting possible future trends for Stablecoins.
COMPUTER STANDARDS & INTERFACES
(2024)
Article
Computer Science, Information Systems
Xin Xu, Guangli Zhu, Houyue Wu, Shunxiang Zhang, Kuan-Ching Li
Summary: As an emotional cause detection task, Emotion-Cause Pair Extraction (ECPE) provides technical support for intelligent psychological counseling, empty-nest elderly care, and other fields. Different from existing methods, this paper proposes an extraction model named SEE-3D, which considers the influence of sentimental intensity to improve extraction accuracy. The results of experiments show that the accuracy of ECPE can be effectively improved by the SEE-3D model.
COMPUTER SCIENCE AND INFORMATION SYSTEMS
(2023)
Review
Computer Science, Information Systems
Shunxiang Zhang, Tong Zhao, Houyue Wu, Guangli Zhu, KuanChing Li
Summary: This paper proposes a classification model for consumer reviews called TS-GCN, which can classify sentiment from both explicit and implicit aspects. The model processes the text using the BERT and BiLSTM models to obtain text features, and the GCN model is used for explicit sentiment classification. The TS model is then used to predict implicit sentiment words, improving the accuracy of sentiment classification.
COMPUTER SCIENCE AND INFORMATION SYSTEMS
(2023)
