检索结果-内蒙古大学图书馆

6th IFIP TC 12 International Conference on Intelligence science, ICIS 2024

作者： Li, Dingyi PCA Lab Key Lab of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education and Jiangsu Key Lab of Image and Video Understanding for Social Security School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing210094 China

ISBN: (纸本)9783031712524

Perceptual video super-resolution aims at converting low-resolution videos to visually appealing high-resolution ones. It may lead to temporal inconsistency due to the drastically changing outputs. In this paper, we propose cascaded sliding-window-based relativistic GAN (Generative Adversarial Network) fusion for perceptual and consistent video super-resolution (PC-VSR). Firstly, cascaded sliding-window-based relativistic GAN is designed to extract more useful information. It enlarges the temporal receptive field of sliding-window-based model in each step. It is able to enhance perceptual quality and compensate temporal consistency progressively and sufficiently. The trained separate refinement generator networks are fused into a final refinement generator. The final refinement generator can be calculated recursively at the testing stage. With our generator fusion, the parameter number is reduced and good quality is maintained. Extensive experimental results demonstrate that our approach outperforms state-of-the-art super-resolution methods in terms of perceptual quality. Our method also achieves good temporal consistency and per-pixel accuracy, compared with other perceptual approaches. © IFIP International Federation for information Processing 2025.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Unformer: A Transformer-Based Approach for Adaptive Multiscale Feature Aggregation in Underwater Image Enhancement

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2025年第4期6卷 1024-1037页

作者： Qing, Yuhao Wang, Yueying Yan, Huaicheng Xie, Xiangpeng Wu, Zhengguang Shanghai University School of Mechatronic Engineering and Automation Shanghai200444 China East China University of Science and Technology School of Information Science and Engineering Shanghai200237 China Nanjing University of Posts and Telecommunications Institute of Advanced Technology Nanjing210023 China Zhejiang University Institute of Cyber-Systems and Control Hangzhou310027 China

Underwater imaging is often compromised by light scattering and absorption, resulting in image degradation and distortion. This manifests as blurred details, color shifts, and diminished illumination and contrast, thereby hindering advancements in underwater research. To mitigate these issues, we propose Unformer, an innovative underwater image enhancement (UIE) technique that leverages a transformer-based architecture for multiscale adaptive feature aggregation. Our approach employs a multiscale feature fusion strategy that adaptively restores illumination and detail features. We reevaluate the relationship between convolution and transformer to develop a novel encoder structure. This structure effectively integrates both long-range and short-range dependencies, dynamically combines local and global features, and constructs a comprehensive global context. Furthermore, we propose a unique multibranch decoder architecture that enhances and efficiently extracts spatial context information through the transformer module. Extensive experiments on three datasets demonstrate that our proposed method outperforms other techniques in both subjective and objective evaluations. Compared with the latest methods, Unformer has improved the peak signal-to-noise ratio (PSNR) by 19.5% and 14.8% respectively on the LSUI and EUVP datasets. The code is available at: https://***/yhflq/Unformer. © 2024 IEEE.

关键词： Light scattering

来源：评论

学校读者我要写书评

暂无评论

High-accuracy face recognition with the MRYOLO-SCO algorithm: combining residual networks and YOLOv2 for improved precision

引用

Signal, Image and Video Processing 2025年第7期19卷 1-12页

作者： Ettiyappan, Anitha palani, Hari kumar Duraipandian, Kavitha Selvaganapathy, Nandhini Department of Information Technology Loyola - ICAM College of Engineering and Technology Nungambakkam Chennai600034 India Department of Computer Science and Engineering MIT School of Computing MIT Art Design and Technology University Loni Kalbhor Pune India Department of computer science and engineering Sathyabama Institute of Science and Technology Chennai India Department of Data Science and Business Systems SRM Institute of Science and Technology Kattankulathur Chengalpattu603203 India

Face recognition in real-world scenarios presents significant challenges due to variations in lighting conditions, occlusions, pose changes, and low-resolution images. To address these challenges, this study proposes a Modified Residual YOLO Network-based Single Candidate Optimization algorithm, which integrates a Modified Residual Network with YOLOv2 to improve recognition accuracy while maintaining computational efficiency. Unlike conventional YOLO models, the proposed framework introduces bipolar feature extraction, enabling the network to capture both high-level global features and low-level local details, thus improving robustness against occlusions and illumination variations. Atailored cost function is introduced to minimize bounding box localization errors and classification inaccuracies, incorporating weighted penalties to enhance model precision. Additionally, a Single Candidate Optimization algorithm is employed to fine-tune hyperparameters, ensuring faster convergence and lower computational complexity. The performance of the proposed technique is evaluated using the benchmark datasets and the simulation results that the proposed framework outperforms existing YOLO-based models, achieving an accuracy of 98.8%, a recall of 98.0%, and an F1-score of 98.1%while maintainingalower computational cost than YOLOv3, YOLOv4, and YOLOv5. The experimental results also indicate that the tailored cost function reduces misclassification errors by 6.1% and accelerates convergence by 20%, making the proposed framework highly suitable for real-time face recognition applications in resource-constrained environmentssuch asIoT-based surveillance and smart security systems. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2025.

关键词： Face recognition

来源：评论

学校读者我要写书评

暂无评论

Power Allocation Based on Federated Multiagent Deep Reinforcement Learning for NOMA Maritime Networks

引用

IEEE Internet of Things Journal 2025年第9期12卷 12869-12884页

作者： He, Jiayi Liu, Zhiyong Zhang, Yakai Jin, Zihao Zhang, Qinyu School of Information Science and Engineering Shandong Weihai264209 China Shandong Provincial Key Laboratory of Marine Electronic Information and Intelligent Unmanned Systems Key Laboratory of Cross-Domain Synergy and Comprehensive Support for Unmanned Marine Systems of the Ministry of Industry and Information Technology Shandong Weihai264209 China Communication Engineering Research Center Guangdong Shenzhen518055 China Peng Cheng Laboratory Shenzhen518055 China

The absence of established communication infrastructures at sea presents a significant hurdle for the development of the Internet of Things (IoT) in maritime environments. In this article, we propose a novel solution: a nonorthogonal multiple access (NOMA) maritime network employing power allocation facilitated by federated multiagent deep reinforcement learning (DRL). Conventional single-agent DRL algorithms encounter challenges, such as dimensionality explosion in real-world scenarios. We address this issue by extending these algorithms to incorporate multiple agents. Furthermore, to mitigate risks associated with centralized training, such as data leakage and network attacks, we adopt a federated learning (FL) framework for distributed training across multiple agents. By uploading only select parameters during training and keeping data locally, FL not only enhances algorithm convergence speed but also bolsters data privacy. Specifically, we introduce the federated multiagent deep Q-network (FLMADQN) algorithm tailored for power allocation in NOMA maritime networks. Our algorithm aims to maximize system throughput, optimize data transfer rates, and expedite convergence. Through extensive computer simulations, we validate the efficacy of our proposed approach. Results demonstrate that the FLMADQN algorithm significantly outperforms the traditional DRL algorithm, DQN, in NOMA maritime environments, improving average system throughput by 20.12%, peak system throughput by 45.10%, and system spectral efficiency by 48.33%. Moreover, FLMADQN exhibits a twofold increase in convergence speed compared to MADQN without FL and is 2.5 times faster than DQN. Our findings underscore the potential of federated multiagent DRL in advancing communication systems for maritime IoT applications. © 2014 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

NeurTV: Total Variation on the Neural Domain*

引用

SIAM Journal on Imaging sciences 2025年第2期18卷 1101-1140页

作者： Luo, Yisi Zhao, Xile Ye, Kai Meng, Deyu School of Mathematics and Statistics Xi'an Jiaotong University Shaanxi Xi'an China School of Automation Science and Engineering Faculty of Electronic and Information Engineering Xi'an Jiaotong University Shaanxi Xi'an China Pengcheng Laboratory Shenzhen China Macao Institute of Systems Engineering Macau University of Science and Technology Taipa China School of Mathematical Sciences University of Electronic Science and Technology of China Sichuan Chengdu China

Recently, we have witnessed the success of total variation (TV) for many imaging applications. However, traditional TV is defined on the original pixel domain, which limits its potential. In this work, we suggest a new TV regularization defined on the neural domain. Concretely, the discrete data is implicitly and continuously represented by a deep neural network (DNN), and we use the derivatives of DNN outputs with respect to (w.r.t.) input coordinates to capture local correlations of data. As compared with classical TV on the original domain, the proposed TV on the neural domain (termed NeurTV) enjoys the following advantages. First, NeurTV is free of discretization error induced by the discrete difference operator. Second, NeurTV is not limited to meshgrid but is suitable for both meshgrid and non-meshgrid data. Third, NeurTV can more exactly capture local correlations across data for any direction and any order of derivatives attributed to the implicit and continuous nature of neural domain. We theoretically reinterpret NeurTV under the variational approximation framework, which allows us to build the connection between NeurTV and classical TV and inspires us to develop variants (e.g., space-variant NeurTV). Extensive numerical experiments with meshgrid data (e.g., color and hyperspectral images) and non-meshgrid data (e.g., point clouds and spatial transcriptomics) showcase the effectiveness of the proposed methods. © 2025 Society for Industrial and Applied Mathematics.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Smart Hybrid Mini-Xception Model and Eyebrow Displacement-Based Stress Level and Emotion Detection System 12th

Smart Hybrid Mini-Xception Model and Eyebrow Displacement-Ba...

引用

12th International Conference on Recent Trends in Computing, ICRTC 2024

作者： Bhuvana, S. Anantha Prabha, P. Deva Priya, M. Christy Jeba Malar, A. Karthick, S. School of Computer Science Engineering and Information Systems Vellore Institute of Technology Vellore India Department of Computer Science and Engineering PSG Institute of Technology and Applied Research Tamilnadu Coimbatore India Department of Computer Science and Engineering Karpagam College of Engineering Tamilnadu Coimbatore India Department of Information Technology Sri Krishna College of Technology Tamilnadu Coimbatore India Department of Computer Science and Engineering Delhi - NCR Campus SRM Institute of Science and Technology Uttar Pradesh Ghaziabad India

ISBN: (纸本)9789819789450

Stress is a state of mental or emotional strain due to adversative or challenging situations. A human may undergo bad life experiences or events, and it is a significant issue to be dealt in today's society. It comes as a result of family situations, health concerns, work environments and people's behavior among other things. It affects people of various ages, including children, teenagers, middle-aged people and senior citizens. It has serious consequences affecting human body and life. These problems can be addressed by detecting mental stress at an early stage. The main goal of the propounded system is to identify human stress using a pre-trained Mini-Xception model. Additional measures such as eye moments, mouth activity and head motion are employed for determining the level of human stress. FER2013 dataset is used and the system is tested for all kinds of emotions. Human stress level is experimented by capturing video using webcam. From the experimental outcomes, it is evident that the proposed system offers better performance in terms of Precision, Accuracy, Specificity and Sensitivity in contrast to traditional methods. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Eyebrow displacement FER2013 dataset Mini-Xception model Stress detection

来源：评论

学校读者我要写书评

暂无评论

TCIP: Transformed Colour Intensity Profiles analysis for fault detection in power line insulators

引用

Multimedia Tools and Applications 2025年第8期84卷 4503-4534页

作者： Tomaszewski, Michal Gasz, Rafal Kasana, Singara Singh Osuchowski, Jakub Singh, Simranjit Zator, Slawomir Department of Computer Science Faculty of Electrical Engineering Automatic Control and Informatics Opole University of Technology Prószkowska 76 Opole45-758 Poland Department of Computer Science and Information Technology Central University of Haryana Haryana Mahendergarh123031 India Department of Information Technology Dr B. R. Ambedkar National Institute of Technology Punjab Jalandhar India Department of Safety Engineering and Technical Systems Faculty of Production Engineering and Logistics Opole University of Technology Prószkowska 76 Opole45-758 Poland

In response to the escalating demand for electricity, the aging process and inherent failures in power lines have become unavoidable challenges in their operational integrity. This research addresses the imperative need for a robust method to detect damage in power insulator disc images, introducing a novel feature extraction technique grounded in color intensity profile transformation and ensemble learning. For the purposes of the research, a specific dataset was developed based on publicly available images of power insulators. The collection consists of 9232 images of insulator discs which included 2184 images of damaged discs. The discernible damages include breakages and burnouts. Two distinct analysis variants were employed. In the first variant (V1), two classification groups were selected: disc with damages and disc without damages. In the second variant (V2), three classification groups were selected: discs without damage, broken discs, and burnout discs. The described research was divided into two successive stages. In the first of them (S1), the transformed color intensity profiles for the prepared data set were built and various machine learning models for their classification were trained and evaluated. In the next stage (S2), previously developed, highest efficient machine learning models, were used to make the detection of various damaged types of overhead line insulator discs. For the first stage of research high-performance results and a high value of the F1 score parameters were obtained (F1 score = 0.999 for the first variant (S1V1) and F1 score = 0.992 for the second variant (S1V2). In the second stage of research and the first variant (S2V1), the highest result of the F1 score was 0.932 and for the second variant (S2V2) F1 score was 0.930. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Prediction and Feedback Assisted Evolutionary Algorithms for Scheduling Urban Traffic Signals

引用

IEEE Transactions on Intelligent Transportation systems 2025年第5期26卷 6881-6890页

作者： Lin, Zhongjie Gao, Kaizhou Duan, Peiyong Wu, Naiqi Suganthan, Ponnuthurai Nagaratnam Macau University of Science and Technology Macau Institute of Systems Engineering Taipa China Zhuhai MUST Science and Technology Research Institute Guangdong Zhuhai519031 China Qilu University of Technology Shandong Academy of Sciences School of Information and Automation Shandong Jinan250102 China Qatar University KINDI Center Doha Qatar

With the acceleration of urbanization, the traffic congestion issue is becoming more and more prominent in large cities. The effective scheduling of urban traffic signals becomes critical. This study proposes three novel prediction and feedback assisted evolutionary algorithms (PFAEAs) to address the urban traffic signal scheduling problem (UTSSP) with minimizing vehicle delays. First, we construct a mathematical model of UTSSP and design an improved evolutionary algorithm (EA) framework that integrates an eight-phase control strategy based on a vehicle movement relationship graph. Then, by combining a back-propagation neural network (BPNN) and meta-heuristics, we improve the prediction accuracy of the vehicle turning rate for generating high-quality initial solutions. Further, 12 problem-specific search operators (PSSOs) are designed to enhance the exploration capability of EA. Reinforcement learning (RL) algorithms, especially the Q-learning and Sarsa algorithms, are employed to select premium PSSOs dynamically for guiding the search direction of EA. Finally, for 18 cases with different scales, the proposed PFAEAs show significant advantages in reducing vehicle delays compared with the state-of-the-art algorithms. The results validate the competitiveness and practicality of the PFAEAs for UTSSP. © 2000-2011 IEEE.

关键词： Traffic congestion

来源：评论

学校读者我要写书评

暂无评论

Copula Entropy Based Causal Network Discovery from Non-stationary Time Series 27th

Copula Entropy Based Causal Network Discovery from Non-stati...

引用

27th International Conference on Pattern Recognition, ICPR 2024

作者： Yang, Jing Rao, Xinzhi The Intelligent Interconnected Systems Hefei University of Technology Hefei230009 China The School of Computer Science and Information Engineering Hefei University of Technology Hefei230009 China

ISBN: (纸本)9783031781827

The task of deducing the causal network from time series data and identifying relationships among multiple series is increasingly vital across various sectors such as industry, medicine, and finance. Despite numerous algorithms being proposed for this purpose, the majority are predicated on the stationarity assumption. However, in disciplines like climatology and neuroscience, time series often exhibit non-stationarity, characterized by a data distribution that shifts over time. In this paper, we introduce an innovative algorithm designed to discern causal relationships from non-stationary time series. Our approach unfolds in three key steps: Initially, we harness the concept of copula entropy to estimate the conditional transfer entropy, offering a streamlined method for non-parametric conditional independence testing. Subsequently, we introduce the time index, which influences other variables at specific time lags, and by integrating the conditional transfer entropy, we execute the independence tests. This leads us to propose the CE-CDN (Copula Entropy-based Causal Discovery from Non-stationary time series), a two-stage algorithm tailored for learning the causal network and identifying change modules. Finally, through comparative analysis with existing algorithms, our experimental findings indicate that CE-CDN not only excels in managing non-stationary time series but also boasts commendable time efficiency. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Time series analysis

来源：评论

学校读者我要写书评

暂无评论

CNN-Based Multi-Output and Multi-Task Regression for Supershape Reconstruction from 3D Point Clouds

Informatica (Slovenia)

引用

Informatica (Slovenia) 2025年第5期49卷 37-48页

作者： Remmach, Hassnae Razak, Siti Fatimah Abdul Ullah, Arif Yogarayan, Sumendra Sayeed, Md Shohel Mrhari, Amine Computer Systems Engineering Laboratory Cadi Ayyad University Marrakesh40000 Morocco Faculty of Information Science and Technology Multimedia University Ayer Keroh75450 Malaysia Research in Computer Science Laboratory Faculty of Sciences Ibn Tofail University Kenitra Morocco

In a number of industries, including computer graphics, robotics, and medical imaging, three-dimensional reconstruction is essential. In this research, a CNN-based Multi-output and Multi-Task Regressor with deep learning capabilities is proposed for three-dimensional object reconstruction from 3D point cloud. Our approach is grounded in the original Point Net architecture, which addresses the difficulties associated with convolution when applied to point clouds. Firstly, this paper is modified using a Multi-Output Regressor to accurately recreate Super forms from 3D point clouds. Using this method, we first extract features from the 3D point cloud using Point Net. After that, a Multi-Output Regressor receives these data and uses them to anticipate the Super shape parameters needed to reconstruct the shape. Taking in the data, the Multi-Output Regressor retrieved characteristics from Point Net and simultaneously predicts several outcomes. Second, a Multi-Task Regressor is used to modify the Point Net. The network gains from the capacity to transfer knowledge from one task to another, improving the model's overall performance. The model would forecast the ten parameters needed to create the shape in the case of rebuilding Super shapes. The test findings were better than expected;they are intriguing in terms of prediction accuracy and cost, and they update the result by 80%, which is a good accomplishment for the study. © 2025 Slovene Society Informatika. All rights reserved.

关键词： 3D reconstruction

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：