Article
Genetics & Heredity
Yanping Li, Nian Fang, Haiquan Wang, Rui Wang
Summary: In this paper, a multi-modal medical image fusion algorithm based on geometric algebra sparse representation is proposed. The algorithm avoids the loss of correlation between channels and outperforms existing methods in subjective and objective quality evaluation.
FRONTIERS IN GENETICS
(2022)
Article
Computer Science, Information Systems
Chengchang Pan, Yongjun Zhang, Zewei Wang, Zhongwei Cui
Summary: In this paper, a robust dictionary learning method based on image multi-vector representation is proposed to balance large-scale information and global features. The method generates a virtual image and obtains multi-vector representation for better image classification accuracy.
Article
Computer Science, Artificial Intelligence
Dan Tang, Qingyu Xiong, Hongpeng Yin, Zhiqin Zhu, Yanxia Li
Summary: This study proposes a sparse representation based image fusion method, which removes unvalued information by constructing a compact dictionary through joint patch grouping and informative sampling. Experimental results demonstrate the superiority of this method over traditional approaches.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Computer Science, Information Systems
Shaoning Zeng, Bob Zhang, Jianping Gou, Yong Xu, Wei Huang
Summary: Dictionary-based classification is effective in knowledge discovery from image data, but faces challenges in balancing the number of dictionary atoms with classification performance, as well as the speed decrease on large datasets. The proposed FRDC framework improves robustness by introducing l(2)-norm optimization and solving optimization based on both l(1)- and l(2)-norms in stages, enhancing robustness, simplicity, and speed.
ACM TRANSACTIONS ON KNOWLEDGE DISCOVERY FROM DATA
(2021)
Article
Computer Science, Artificial Intelligence
Arash Abdi, Mohammad Rahmati, Mohammad M. Ebadzadeh
Summary: In this paper, a new discriminative dictionary learning algorithm is proposed, which embeds an entropy-based criterion in the objective function to enforce a proper structure for dictionary items. Experimental results demonstrate that the algorithm outperforms other methods on various real-world image datasets.
PATTERN RECOGNITION
(2021)
Article
Computer Science, Artificial Intelligence
Chang Wang, Yang Wu, Yi Yu, Jun Qiang Zhao
Summary: In this study, an improved multi-modality image fusion method was proposed by combining the joint patch clustering-based adaptive dictionary and sparse representation to address the issue of gray inconsistency caused by the maximum L-1 norm fusion rule. Through quantitative evaluation and comparative experiments, it was demonstrated that the method has superiority in fusion metrics, image quality, and edge preservation.
MACHINE VISION AND APPLICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Sumit Budhiraja, Rajat Sharma, Sunil Agrawal, Balwinder S. Sohi
Summary: An efficient image fusion method based on sparse representation with clustered dictionary is proposed in this paper for infrared and visible images. By enhancing the edge information of visible image using a guided filter and using non-subsampled contourlet transform for fusion, the proposed method is able to outperform other conventional image fusion methods.
PATTERN ANALYSIS AND APPLICATIONS
(2021)
Article
Engineering, Electrical & Electronic
Jin Tan, Taiping Zhang, Linchang Zhao, Xiaoliu Luo, Yuan Yan Tang
Summary: This study introduces a novel multi-focus image fusion algorithm based on Geometrical Sparse Representation (GSR). The algorithm uses a single dictionary image and a weighted GSR model to assess the focus activity level of source images, refines the decision map through simple post-processing, and ultimately generates an all-in-focus image.
SIGNAL PROCESSING-IMAGE COMMUNICATION
(2021)
Article
Engineering, Electrical & Electronic
Wenqing Wang, Xiao Ma, Han Liu, Yuxing Li, Wei Liu
Summary: This study introduces a novel fusion method based on joint convolutional analysis and synthesis (JCAS) sparse representation, which can effectively reduce spatial artifacts and blurring effects in multi-focus image fusion, producing clearer edge details.
SIGNAL PROCESSING-IMAGE COMMUNICATION
(2021)
Article
Computer Science, Artificial Intelligence
Usman Haider, Muhammad Hanif, Ahmar Rashid, Syed Fawad Hussain
Summary: This paper proposes a dictionary-based training method for ConvNets that reduces training time significantly while maintaining accuracy by exploiting redundancy in the training data. Experimental results on three publicly available datasets show a 4.5 times reduction in computational burden compared to state-of-the-art algorithms like ResNet-{18,34,50}, with comparable accuracy.
IMAGE AND VISION COMPUTING
(2023)
Article
Computer Science, Information Systems
Ali Nozaripour, Hadi Soltanizadeh
Summary: Convolutional Sparse Coding (CSC) is a popular model in signal and image processing, and this paper proposes a novel discriminative model based on CSC for image classification. Experimental results demonstrate the superior performance of the proposed method.
MULTIMEDIA TOOLS AND APPLICATIONS
(2022)
Article
Physics, Multidisciplinary
Bingzhe Wei, Xiangchu Feng, Kun Wang, Bian Gao
Summary: A novel fusion method that combines CNN and SR for multi-focus image fusion has been proposed, resulting in a more accurate and informative fused image. Experimental results demonstrate that this method clearly outperforms existing methods in terms of visual perception and objective evaluation metrics, while also significantly reducing computational complexity.
Article
Computer Science, Information Systems
Xiaole Ma, Zhihai Wang, Shaohai Hu
Summary: A fusion method based on multi-scale sparse representation for registered multi-focus images (MIF-MsSR) is proposed in this paper, with an adaptive fusion rule for sparse coefficients presented. Experiments have shown that the proposed method not only preserves the integrity of information in source images, but also outperforms other state-of-the-art methods in fusion performance on subjective and objective indicators.
JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION
(2021)
Article
Engineering, Electrical & Electronic
Minghong Xie, Jiaxin Wang, Yafei Zhang
Summary: This paper presents a unified framework for image fusion and completion, achieving separation and restoration of different components through a low-rank and sparse dictionary learning model to recover lost information of damaged images. The maximum l(1)-norm fusion scheme is adopted to merge coding coefficients of different components. Experimental results demonstrate that this method excels in preserving image brightness and details, outperforming other methods in performance evaluation.
SIGNAL PROCESSING-IMAGE COMMUNICATION
(2021)
Article
Computer Science, Artificial Intelligence
Majid Ghasemi, Manoochehr Kelarestaghi, Farshad Eshghi, Arash Sharifi
Summary: The study introduces an adaptive fuzzy dictionary learning method for image classification, which optimizes basis vectors to accurately represent data and addresses the inherent uncertainty in input data. Experimental results demonstrate the superior performance of the method in medical image classification, with high accuracy, sensitivity, and specificity.
PATTERN ANALYSIS AND APPLICATIONS
(2021)
Article
Computer Science, Software Engineering
Yuanyuan Li, Ziyu Wang, Li Yin, Zhiqin Zhu, Guanqiu Qi, Yu Liu
Summary: This paper proposes a dual encoding-decoding structure of X-shaped network (X-Net) that integrates the characteristics of CNNs and Transformer. It can serve as a good alternative to the traditional pure convolutional medical image segmentation network.
Article
Automation & Control Systems
Xiaoyu Fang, Jianfeng Qu, Yi Chai, Bowen Liu
Summary: This paper proposes an adaptive multiscale and dual subnet convolutional auto-encoder (AMDSCAE) to detect intermittent faults (IFs) in analog circuits. The AMDSCAE method can assign different attention and fuse multiscale information adaptively, resulting in better noise robustness. Additionally, the dual subnet structure enhances the IF detection ability and can detect weaker faults. Experimental results on three typical analog filter circuits demonstrate that AMDSCAE has better noise immunity and can detect weaker IFs.
Article
Computer Science, Artificial Intelligence
Yiyao An, Ke Zhang, Yi Chai, Qie Liu, Xinghua Huang
Summary: In this paper, a domain adaptation network based on contrastive learning (DACL) is proposed for bearing fault diagnosis. The method consists of a feature mining module and an adversarial domain adaptation module, addressing the issues of similarity in fault features and misclassification near the distribution boundaries. Experimental results demonstrate the effectiveness of the proposed method in various fault diagnosis scenarios.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Theory & Methods
Shujuan Wang, Run Liu, Huafeng Li, Guanqiu Qi, Zhengtao Yu
Summary: Due to incomplete appearance features, matching occluded pedestrians under multiple cross-camera views is a long-term challenge. This paper introduces the idea of adversarial attack into occluded person re-ID and proposes an adversarial training framework to defend against obstacles and improve pedestrian identity matching. The proposed framework broadens research horizons in robust model design and achieves better performance on occluded re-ID datasets.
IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY
(2023)
Letter
Computer Science, Information Systems
Yingming Tian, Yi Chai, Li Feng, Ke Zhang
SCIENCE CHINA-INFORMATION SCIENCES
(2023)
Letter
Computer Science, Information Systems
Zheren Zhu, Pengfei Huang, Xinmin Zhang, Yi Chai, Zhihuan Song
SCIENCE CHINA-INFORMATION SCIENCES
(2023)
Article
Physics, Multidisciplinary
Yuhu Liu, Xiaolong Chen, Yongfang Mao, Yi Chai, Yutao Jiang
Summary: This study proposes a novel method that combines an improved energy fluctuation index (IEFI) and modified VMD (MVMD) to solve the issues of mode number and balance parameter in VMD application. Simulation results show that the proposed method can effectively detect the feature of a periodic impulse signal and accurately diagnose bearing faults.
FRONTIERS IN PHYSICS
(2023)
Article
Engineering, Electrical & Electronic
Xiaoyu Fang, Jianfeng Qu, Qiu Tang, Yi Chai
Summary: Intermittent faults (IFs) are short-lived and repeatable failures in analog circuits, which are difficult to label. Unsupervised cluster recognition is an important method to analyze potential classes of IFs. However, outliers from noise can impact the IF recognition. This article proposes an improved density peak clustering (DPC) method that enhances fault recognition by improving distance measure and outlier detection.
IEEE SENSORS JOURNAL
(2023)
Article
Automation & Control Systems
Xiaolong Chen, Yi Chai, Qie Liu, Pengfei Huang, Linchuan Fan
Summary: In this paper, a novel Bayesian sparse multiple kernel-based identification method (BSMKM) for multiple-input single-output (MISO) Hammerstein system is proposed. The method represents the nonlinear part and the linear part using basis-function model and finite impulse response model respectively and estimates all unknown model parameters through hierarchical prior distribution and full Bayesian method based on variational Bayesian inference.
Article
Engineering, Industrial
Xiaoyu Fang, Jianfeng Qu, Yi Chai
Summary: This paper proposes a prior knowledge-guided teacher-student (PKGTS) model based on self-supervised learning to improve fault detection. The model introduces prior knowledge of intermittent faults through pretext tasks and achieves IF detection through the cognitive biases of faults.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2023)
Article
Engineering, Industrial
Han Zhou, Hongpeng Yin, Yi Chai
Summary: This paper proposes a novel method for multimode industrial processes fault diagnosis, which utilizes hierarchical clustering strategy to analyze the multi-grained information of process data and model the correlations between operating modes and patterns within each mode. A feature learning algorithm based on nonnegative matrix factorization (NMF) is then proposed to learn data features and represent samples by discovered multi-grained structural information. Moreover, a weighted metric is designed to measure the feature similarities learned by NMF. The effectiveness of the proposed framework is validated on a numerical example and a multiple-phase flow process.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2023)
Article
Automation & Control Systems
Qie Liu, Biao Huang, Yi Chai, Wenbo Li
Summary: The identification of topology in sparse networks is crucial for network modeling in various fields. An efficient algorithm based on stochastic optimization is proposed to decrease computational complexity and is suitable for network identification with large data sets.
Article
Computer Science, Artificial Intelligence
Linchuan Fan, Yi Chai, Xiaolong Chen
Summary: MEB is a performance enhancement framework that employs multi-scale ensemble booster to help existing TSD classifiers achieve performance leaps, significantly improving model performance. The framework utilizes an easy-to-combine network structure and a probability distribution co-evolution strategy to optimize label probability distribution.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Engineering, Electrical & Electronic
Shuiqing Xu, Wenzhan Huang, Darong Huang, Hongtian Chen, Yi Chai, Mingyao Ma, Wei Xing Zheng
Summary: This article presents a reduced-order observer-based simultaneous diagnosis strategy for grid-tied neutral point clamped (NPC) inverters subjected to open-switch and current sensor faults. The strategy involves constructing an augmented descriptor system to transfer the current sensor fault into a generalized state vector, applying matrix transformations to decouple the open-switch fault from the inverter system state and the current sensor fault, developing a reduced-order observer for precise estimation of the phase current and the current sensor fault, and proposing a diagnosis algorithm with an adaptive threshold for distinguishing between different types of faults and locating the faulty power switch. Experimental results and comparisons confirm the effectiveness of the proposed fault diagnosis algorithm.
IEEE TRANSACTIONS ON POWER ELECTRONICS
(2023)
Article
Biology
Zhiqin Zhu, Zheng Yao, Xin Zheng, Guanqiu Qi, Yuanyuan Li, Neal Mazur, Xinbo Gao, Yifei Gong, Baisen Cong
Summary: Drug-target affinity (DTA) prediction is an emerging and effective method in drug development research to evaluate the efficacy and safety of candidate drugs. However, existing DTA prediction models lack information on interactions between molecular substructures, impacting prediction accuracy and interpretability. Therefore, TDGraphDTA is introduced, using Transformer and Diffusion to predict drug-target interactions by incorporating multi-scale information interaction and graph optimization.
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)
