检索结果-内蒙古大学图书馆

Blockchain-based knowledge-aware semantic communications for remote driving image transmission

Digital Communications and Networks 2025年第2期11卷 317-325页

作者： Yangfei Lin Tutomu Murase Yusheng Ji Wugedele Bao Lei Zhong Jie Li Department of Computer and Network Engineering University of Electro-Communications Information Technology Center Nagoya University National Institute of Informatics(NII) Hohhot Minzu College Toyota Motor Corporation Department of Computer Science and Engineering Shanghai Jiaotong University

Remote driving, an emergent technology enabling remote operations of vehicles, presents a significant challenge in transmitting large volumes of image data to a central server. This requirement outpaces the capacity of traditional communication methods. To tackle this, we propose a novel framework using semantic communications, through a region of interest semantic segmentation method, to reduce the communication costs by transmitting meaningful semantic information rather than bit-wise data. To solve the knowledge base inconsistencies inherent in semantic communications, we introduce a blockchain-based edge-assisted system for managing diverse and geographically varied semantic segmentation knowledge bases. This system not only ensures the security of data through the tamper-resistant nature of blockchain but also leverages edge computing for efficient management. Additionally, the implementation of blockchain sharding handles differentiated knowledge bases for various tasks, thus boosting overall blockchain efficiency. Experimental results show a great reduction in latency by sharding and an increase in model accuracy, confirming our framework's effectiveness.

关键词： Semantic communication Remote driving Semantic segmentation Blockchain Knowledge base management

来源：评论

学校读者我要写书评

暂无评论

A novel dominating set and centrality based graph convolutional network for node classification

引用

Multimedia Tools and Applications 2025年第14期84卷 12979-13004页

作者： Garg, Neeraj Garg, Sneha Jadeja, Mahipal Meena, Yogesh Kumar Gopalani, Dinesh Chauhan, Ganpat Singh Department of Computer Science and Engineering Malaviya National Institute of Technology Jaipur Rajasthan Jaipur302017 India Department of Computer Science and Engineering Swami Keshvanand Institute of Technology Jaipur Rajasthan Jaipur302017 India Department of Information Technology School of IT Manipal University Jaipur Rajasthan Jaipur302034 India

To learn and analyze graph-structured data, Graph Neural Networks (GNNs) have emerged as a powerful framework over traditional neural networks, which work well on grid-like or sequential structure data. GNNs are particularly designed to record dependencies and relations in large and complicated networks such as social networks, molecular structures, citation networks and recommendation systems. The main idea behind GNN is the message-passing mechanism that allows sharing of information between nodes and edges in a graph network, enabling the model to learn about graph structures and perform well even with limited training data. GNNs have three different architectures: Graph Convolutional Network (GCN), GraphSAGE, Graph Attention Model (GAT) and other hybrid models. In this paper, we have proposed novel GCN models namely dominating Set with Degree Centrality based GCN (DS_DC_GCN) and dominating Set with Betweenness Centrality based GCN (DS_BWC_GCN) model, for node classification task. In the traditional GNN models, training nodes are selected randomly. In our proposed models, we first identify a subset of the nodes with the help of a dominating set (D) of the underlying network. Then we consider final training nodes (say D′) from set D based on the centrality scores (degree or betweenness). We give preference to nodes having high centrality scores for training. To evaluate the performance of the proposed models, we have used three standard citation network datasets - Cora, Citeseer and Pubmed. Our models obtain improved node classification performance as compared to other state-of-the-art-methods. The consideration of high centrality nodes of dominating set in training helps to improve the feature availability in the underlying computational graphs, which in turn improves learning and hence, model performance. These models also have the potential to offer better results for other network tasks, including link prediction and community detection. © The Author(s), under

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

PerFedKG: two-stage information-loop federated knowledge graph for personalized privacy-preserving recommendation systems

引用

science China(information sciences) 2025年第4期68卷 98-99页

作者： Fan WANG Xuyun ZHANG Weiming LIU Li LI Yuwen LIU Zhongyuan ZHANG Guanfeng LIU Shengye PANG Xiaolong XU Lianyong QI College of Computer Science and Technology Zhejiang University Department of Computing Macquarie University College of Computer Science and Technology China University of Petroleum (East China) School of Software Nanjing University of Information Science and Technology

Knowledge graphs(KGs) effectively mitigate data sparsity in recommendation systems(RSs) by providing valuable auxiliary information [1]. However, traditional centralized KG-based RSs increase the risk of user privacy *** learning(FL) enhances RS's privacy by enabling model training on decentralized data [2]. Although integrating KG and FL can address both data sparsity and privacy issues in RSs [3], several challenges persist. CH1,Each client's local model relies on a consistent global model from the server, limiting personalized deployment to endusers.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Absorbing Time Synchronization Errors Using Shadow Queues in Time-Sensitive Networking

引用

IEEE Internet of Things Journal 2025年第9期12卷 12912-12926页

作者： Yang, Hao Zhang, Tong Feng, Xiaoqin Ma, Yanying Zhang, Liwei Ren, Fengyuan Lanzhou University School of Information Science and Engineering Lanzhou730000 China Nanjing University of Aeronautics and Astronautics College of Computer Science and Technology Nanjing210016 China Tsinghua University Department of Computer Science and Technology Beijing100084 China

Time-sensitive networking (TSN) is widely used in industrial automation and automotive applications due to its ability to provide deterministic transmission. To meet the stringent deterministic requirements of time-sensitive traffic, the mainstream traffic management mechanism time-aware shaper (TAS) relies on time synchronization among network devices. TAS schedules periodic traffic transmission using preallocated time windows. However, in real networks, time-sensitive traffic may not be precisely forwarded as scheduled, thus failing to achieve the expected performance. A fundamental reason is that the statically planned transmission schemes cannot accommodate dynamic time synchronization errors. To address this problem, we propose a novel TSN traffic scheduling strategy called robust TAS (RTAS), which utilizes a dual-queue gating structure to define a new scheduling rule for absorbing time synchronization errors to guarantee the real-time performance of scheduled traffic (ST). We conducted extensive simulations on OMNeT++ to study the performance of RTAS in industrial automation scenarios. The results show that RTAS enables deterministic transmission with imperfectly synchronized clocks and minimizes the impact of time synchronization errors on ST. © 2014 IEEE.

关键词： Queueing theory

来源：评论

学校读者我要写书评

暂无评论

Enhancing Accuracy of Handwritten Character Segmentation Using Region-Based Convolutional Neural Networks (R-CNN) 1

Enhancing Accuracy of Handwritten Character Segmentation Usi...

引用

1st International Conference on AIML-Applications for engineering and technology, ICAET 2025

作者： Varpe, Kanchan Sakhare, Sachin Joshi, Rahul Vishwakarma Institute of Information Technology Department of Computer Engineering Pune India Department of Computer Science and Engineering Pune India

ISBN: (纸本)9798350355611

Handwritten character segmentation plays a pivotal role in the performance of Optical Character Recognition (OCR) systems. This paper introduces an innovative approach to enhancing segmentation accuracy using Region-Based Convolutional Neural Networks (R-CNN). Traditional methods often struggle with issues such as varying character sizes, overlapping, and inconsistent writing styles, leading to suboptimal recognition results. The proposed R-CNN model addresses these challenges by effectively localizing and segmenting individual characters from handwritten text images. The methodology involves training the R-CNN on a comprehensive dataset of handwritten characters, annotated to capture diverse styles and complexities. The model leverages deep learning's feature extraction capabilities and region-based localization to accurately segment characters, even in challenging scenarios like connected or touching characters. Comparative experiments show that the proposed approach outperforms conventional techniques in terms of segmentation accuracy and robustness. Results demonstrate significant improvements, with the R-CNN model achieving higher precision and recall rates across multiple handwritten datasets, including different languages and scripts. This advancement holds promise for enhancing OCR systems used in applications such as automated document processing and digital archiving. Future work will focus on integrating this segmentation method with advanced recognition models to further boost overall handwritten text recognition performance. © 2025 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

YOLO-LFD: A Lightweight and Fast Model for Forest Fire Detection

引用

computers, Materials & Continua 2025年第2期82卷 3399-3417页

作者： Honglin Wang Yangyang Zhang Cheng Zhu School of Artificial Intelligence Nanjing University of Information Science and TechnologyNanjing210044China School of Computer Science Nanjing University of Information Science and TechnologyNanjing210044China Electrical&Computer Engineering University of Illinois at Urbana-ChampaignUrbanaIL 61801USA

Forest fires pose a serious threat to ecological balance, air quality, and the safety of both humans and wildlife. This paper presents an improved model based on You Only Look Once version 5 (YOLOv5), named YOLO Lightweight Fire Detector (YOLO-LFD), to address the limitations of traditional sensor-based fire detection methods in terms of real-time performance and accuracy. The proposed model is designed to enhance inference speed while maintaining high detection accuracy on resource-constrained devices such as drones and embedded systems. Firstly, we introduce Depthwise Separable Convolutions (DSConv) to reduce the complexity of the feature extraction network. Secondly, we design and implement the Lightweight Faster Implementation of Cross Stage Partial (CSP) Bottleneck with 2 Convolutions (C2f-Light) and the CSP Structure with 3 Compact Inverted Blocks (C3CIB) modules to replace the traditional C3 modules. This optimization enhances deep feature extraction and semantic information processing, thereby significantly increasing inference speed. To enhance the detection capability for small fires, the model employs a Normalized Wasserstein Distance (NWD) loss function, which effectively reduces the missed detection rate and improves the accuracy of detecting small fire sources. Experimental results demonstrate that compared to the baseline YOLOv5s model, the YOLO-LFD model not only increases inference speed by 19.3% but also significantly improves the detection accuracy for small fire targets, with only a 1.6% reduction in overall mean average precision (mAP)@0.5. Through these innovative improvements to YOLOv5s, the YOLO-LFD model achieves a balance between speed and accuracy, making it particularly suitable for real-time detection tasks on mobile and embedded devices.

关键词： Forest fire detection YOLOv5 lightweight small object detection

来源：评论

学校读者我要写书评

暂无评论

A Semantic-Based Approach for Self-Configuring Smart Home Systems Using Feedback-Based Reinforcement Learning

引用

IEEE Internet of Things Journal 2025年第11期12卷 17729-17741页

作者： Ran, Weizhi Nantogma, Sulemana Yang, QingZhuo Zhang, Shangyan Wang, Yan Xu, Yang University of Electronic Science and Technology of China School of Computer Science and Engineering Chengdu611731 China Xi'an University of Architecture and Technology School of Information and Control Engineering China

As urban populations grow, smart home technology has become a key enabler for enhancing energy efficiency, comfort, and convenience in residential environments. However, existing smart home implementations often struggle to adapt to complex and dynamic user needs due to issues like device heterogeneity, rigid pre-set solutions, and limited interoperability. This paper presents a novel semantic-based self-configuring framework that captures the functional, operational, state, and characteristic perspectives of smart home devices, facilitating seamless interconnection and interoperability across heterogeneous devices. The proposed framework uses service entity, that is semantically modeled to act as a linker and mediator between applications and devices to enable dynamic scheduling and control based on application requests and user preferences. To enable continuous adaptation and learning, the proposed approach formulates the self-configuration problem as a Markov Decision Process (MDP), with parameters instantiated using semantic information and rules. A novel Feedback-based Reinforcement Learning approach is then proposed to adjust the reward function and state transition parameters based on user feedback. The performances of the proposed approach against a bench-marked fixed and rule-based mechanisms are demonstrated through simulation. © 2014 IEEE.

关键词： Smart homes

来源：评论

学校读者我要写书评

暂无评论

Systematic Selection of a Suitable Data Imputation Technique Based on Data Characteristics

引用

SN computer science 2025年第5期6卷 1-19页

作者： Sebastian, Anu Maria Peter, David Artificial Intelligence Lab Department of Computer Science Cochin University of Science and Technology Kerala Cochin India Department of Computer Science and Engineering Indian Institute of Information Technology Kottayam India

Analyzing incomplete data is one of the prime concerns in data analysis. Discarding the missing records or values might result in inaccurate analysis outcomes or loss of helpful information, especially when the size of the data is small. A preferable alternative is to substitute the missing values using imputation such that the substituted values are very close to the actual missing values and this is a challenging task. In spite of the existence of many imputation algorithms, there is no universal imputation algorithm that can yield the best values for imputing all types of datasets. This is mainly because of the dependence of the imputation algorithm on the inherent properties of the data. These properties include type of data distribution, data size, dimensionality, presence of outliers, data dependency among the attributes, and so on. In the literature, there exists no straightforward method for determining a suitable imputation algorithm based on the data characteristics. The existing practice is to conduct exhaustive experimentation using the available imputation techniques with every dataset and this requires a lot of time and effort. Moreover, the current approaches for checking the suitability of imputations cannot be done when the ground truth data is not available. In this paper, we propose a new method for the systematic selection of a suitable imputation algorithm based on the inherent properties of the dataset which eliminates the need for exhaustive experimentation. Our method determines the imputation technique which consistently gives lower errors while imputing datasets with specific properties. Also, our method is particularly useful when the real-world data do not have the ground truth for missing data to check the imputation performance and suitability. Once the suitability of a DI technique is established based on the data properties, this selection will remain valid for another dataset with similar properties. Thus, our method can save time an

关键词： Data analysis Data distribution Data imputation Data preprocessing Imputation performance Machine learning Missing value handling

来源：评论

学校读者我要写书评

暂无评论

NJmat 2.0:User Instructions of Data-Driven Machine Learning Interface for Materials science

引用

computers, Materials & Continua 2025年第4期83卷 1-11页

作者： Lei Zhang Hangyuan Deng Department of Internet Engineering School of Software EngineeringNanjing University of Information Science&TechnologyNanjing210044China Department of Materials Physics School of Chemistry and Materials ScienceNanjing University of Information Science&TechnologyNanjing210044China

NJmat is a user-friendly,data-driven machine learning interface designed for materials design and *** platform integrates advanced computational techniques,including natural language processing(NLP),large language models(LLM),machine learning potentials(MLP),and graph neural networks(GNN),to facili-tate materials *** platform has been applied in diverse materials research areas,including perovskite surface design,catalyst discovery,battery materials screening,structural alloy design,and molecular *** automating feature selection,predictive modeling,and result interpretation,NJmat accelerates the development of high-performance materials across energy storage,conversion,and structural ***,NJmat serves as an educational tool,allowing students and researchers to apply machine learning techniques in materials science with minimal coding *** automated feature extraction,genetic algorithms,and interpretable machine learning models,NJmat simplifies the workflow for materials informatics,bridging the gap between AI and experimental materials *** latest version(available at https://***/articles/software/NJmatML/24607893(accessed on 01 January 2025))enhances its functionality by incorporating NJmatNLP,a module leveraging language models like MatBERT and those based on Word2Vec to support materials prediction *** utilizing clustering and cosine similarity analysis with UMAP visualization,NJmat enables intuitive exploration of materials *** NJmat primarily focuses on structure-property relationships and the discovery of novel chemistries,it can also assist in optimizing processing conditions when relevant parameters are included in the training *** providing an accessible,integrated environment for machine learning-driven materials discovery,NJmat aligns with the objectives of the Materials Genome Initiative and promotes broader adoption of AI techniques in materials science.

关键词： Data-driven machine learning natural language processing machine learning potential large language model

来源：评论

学校读者我要写书评

暂无评论

Gloss-driven Conditional Diffusion Models for Sign Language Production

引用

ACM Transactions on Multimedia Computing, Communications and Applications 2025年第4期21卷 1-17页

作者： Tang, Shengeng Xue, Feng Wu, Jingjing Wang, Shuo Hong, Richang School of Computer Science and Information Engineering Hefei University of Technology Hefei China School of Software Hefei University of Technology Hefei China School of Data Science School of Information Science and Technology University of Science and Technology of China Hefei China

Sign Language Production (SLP) aims to convert text or audio sentences into sign language videos corresponding to their semantics, which is challenging due to the diversity and complexity of sign languages, and cross-modal semantic mapping issues. In this work, we propose a Gloss-driven Conditional Diffusion Model (GCDM) for SLP. The core of the GCDM is a diffusion model architecture, in which the sign gloss sequence is encoded by a Transformer-based encoder and input into the diffusion model as a semantic prior condition. In the process of sign pose generation, the textual semantic priors carried in the encoded gloss features are integrated into the embedded Gaussian noise via cross-attention. Subsequently, the model converts the fused features into sign language pose sequences through T-round denoising steps. During the training process, the model uses the ground-truth labels of sign poses as the starting point, generates Gaussian noise through T rounds of noise, and then performs T rounds of denoising to approximate the real sign language gestures. The entire process is constrained by the MAE loss function to ensure that the generated sign language gestures are as close as possible to the real labels. In the inference phase, the model directly randomly samples a set of Gaussian noise, generates multiple sign language gesture sequence hypotheses under the guidance of the gloss sequence, and outputs a high-confidence sign language gesture video by averaging multiple hypotheses. Experimental results on the Phoenix2014T dataset show that the proposed GCDM method achieves competitiveness in both quantitative performance and qualitative visualization. © 2025 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Forming

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：