检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Yao, Dezhong Shi, Yuexin Liu, Tongtong Xu, Zhiqiang National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China Mohamed bin Zayed University of Artificial Intelligence United Arab Emirates

Federated Learning (FL) is increasingly adopted in edge computing scenarios, where a large number of heterogeneous clients operate under constrained or sufficient resources. The iterative training process in conventional FL introduces significant computation and communication overhead, which is unfriendly for resource-constrained edge devices. One-shot FL has emerged as a promising approach to mitigate communication overhead, and model-heterogeneous FL solves the problem of diverse computing resources across clients. However, existing methods face challenges in effectively managing model-heterogeneous one-shot FL, often leading to unsatisfactory global model performance or reliance on auxiliary datasets. To address these challenges, we propose a novel FL framework named FedMHO, which leverages deep classification models on resource-sufficient clients and lightweight generative models on resource-constrained devices. On the server side, FedMHO involves a two-stage process that includes data generation and knowledge fusion. Furthermore, we introduce FedMHO-MD and FedMHO-SD to mitigate the knowledge-forgetting problem during the knowledge fusion stage, and an unsupervised data optimization solution to improve the quality of synthetic samples. Comprehensive experiments demonstrate the effectiveness of our methods, as they outperform state-of-the-art baselines in various experimental setups. Our code is available at https://***/YXShi2000/FedMHO. © 2025, CC BY.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Generalization-Enhanced Code Vulnerability Detection via Multi-Task Instruction Fine-Tuning

arXiv

引用

arXiv 2024年

作者： Du, Xiaohu Wen, Ming Zhu, Jiahao Xie, Zifan Ji, Bin Liu, Huijun Shi, Xuanhua Jin, Hai School of Cyber Science and Engineering Huazhong University of Science and Technology China School of Computer Science and Technology Huazhong University of Science and Technology China College of Computer National University of Defense Technology China National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab HUST Wuhan430074 China Hubei Engineering Research Center on Big Data Security Hubei Key Laboratory of Distributed System Security HUST Wuhan430074 China JinYinHu Laboratory Wuhan430077 China Cluster and Grid Computing Lab HUST Wuhan430074 China

Code Pre-trained Models (CodePTMs) based vulnerability detection have achieved promising results over recent years. However, these models struggle to generalize as they typically learn superficial mapping from source code to labels instead of understanding the root causes of code vulnerabilities, resulting in poor performance in real-world scenarios beyond the training instances. To tackle this challenge, we introduce VulLLM, a novel framework that integrates multi-task learning with Large Language Models (LLMs) to effectively mine deep-seated vulnerability features. Specifically, we construct two auxiliary tasks beyond the vulnerability detection task. First, we utilize the vulnerability patches to construct a vulnerability localization task. Second, based on the vulnerability features extracted from patches, we leverage GPT-4 to construct a vulnerability interpretation task. VulLLM innovatively augments vulnerability classification by leveraging generative LLMs to understand complex vulnerability patterns, thus compelling the model to capture the root causes of vulnerabilities rather than overfitting to spurious features of a single task. The experiments conducted on six large datasets demonstrate that VulLLM surpasses seven state-of-the-art models in terms of effectiveness, generalization, and robustness. © 2024, CC BY.

关键词： Large datasets

来源：评论

学校读者我要写书评

暂无评论

Efficient Node Reduction Heuristic for GNN-Based Traffic Speed Forecasting

Efficient Node Reduction Heuristic for GNN-Based Traffic Spe...

引用

IEEE/ACM International Symposium on Big Data Computing (BDC)

作者： Yuto Inokuchi Pedro Henrique González Masayoshi Aritsugi Israel Mendonça Graduate School of Science and Technology Kumamoto University Kumamoto Japan System Engineering and Computer Science Program Federal University of Rio de Janeiro Rio de Janeiro Brazil Faculty of Advanced Science and Technology Kumamoto University Kumamoto Japan

ISBN: (数字)9798350367300

ISBN: (纸本)9798350367317

In this paper, we aim to reduce the number of nodes from Graph Neural Networks (GNNs), thereby simplifying models and reducing computational costs. GNNs are highly effective for various tasks, such as prediction, classification, and clustering, due to their ability to learn node and edge attributes and relationships, and they have been utilized for intelligent transportation systems recently by converting sensor networks into graph structures. Deep spatio-temporal neural networks, including Spatio-Temporal Graph Convolutional Networks (STGCNs), capture spatial and temporal dependencies, making them suitable for traffic speed forecasting, traffic demand prediction, and travel time estimation. Despite their success, GNNs face challenges in industrial applications due to significant memory usage and time consumption. In this paper, we propose a new approach to node reduction that outperforms existing methods in computational efficiency. Our experiments on two real-world traffic datasets demonstrate that using the heuristic and edge information to reduce nodes can cut computation time of optimization up to 95% and, by eliminating noise, can even enhance prediction accuracy.

关键词： Graph convolutional networks Noise Memory management Estimation Computational efficiency Forecasting Optimization Intelligent transportation systems Faces Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

A Lightweight Multi-Layer Perceptron for Efficient Multivariate Time Series Forecasting

SSRN

引用

SSRN 2023年

作者： Wang, Zhenghong Ruan, Sijie Huang, Tianqiang Zhou, Haoyi Zhang, Shanghang Wang, Yi Wang, Leye Huang, Zhou Liu, Yu Institute of Remote Sensing and Geographic Information System School of Earth and Space Sciences Peking University Beijing China School of Computer Science and Technology Beijing Institute of Technology Beijing China College of Computer and Cyber Security Fujian Normal University Fujian China School of Computer Science and Engineering Beihang University Beijing China School of Computer Science Peking University Beijing China

Efficient and effective multivariate time series (MTS) forecasting is critical for real-world applications, such as traffic management and energy dispatching. Most of the current deep learning studies (e.g., Spatio-Temporal Graph Neural Networks and Transformers) fall short in a trade-off between performance and efficiency. Existing MTS forecasting studies have yet to fully and simultaneously address issues such as modelling both temporal and variate dependencies, as well as the temporal locality, hindering broader applications. In this paper, we propose a lightweight model, i.e., Time Series MLP (TSP). TSP is built upon MLP and relies on the PrecMLP with the proposed computationally free Precurrent mechanism to model both the variate dependency and temporal locality, thus being simple, effective and versatile. Extensive experiments show that TSP outperforms state-of-the-art methods on 16 datasets for both Long-term Time-series Forecasting and Traffic Forecasting tasks. Furthermore, it attains a significant reduction of at least 95.97% in practical training speed on the CPU. © 2023, The Authors. All rights reserved.

关键词： Multilayer neural networks

来源：评论

学校读者我要写书评

暂无评论

Driver Fatigue Detection Using OpenCV and Dlib Library

Driver Fatigue Detection Using OpenCV and Dlib Library

引用

2023 IEEE Asia-Pacific Conference on computer science and Data engineering, CSDE 2023

作者： Shen, Kuang Ramli, Rusyaizila Chuah, Joong Huang Chai, Grace Mei Ting School of Electronics and Computer Science University of Southampton Southampton United Kingdom University of Southampton Malaysia Carbon Neutrality Research Group Johor Iskandar Puteri79100 Malaysia University of Malaya Faculty of Engineering Department of Electrical Engineering Kuala Lumpur50603 Malaysia University of Southampton Malaysia Smart Manufacturing System Reasearch Group Johor Iskandar Puteri79100 Malaysia

ISBN: (纸本)9798350341072

Driver fatigue and distraction are major contributors to fatal road accidents worldwide. To enhance vehicle safety, this article proposes a fatigue detection system using EAR and MAR algorithms to identify signs of fatigue, such as eye closure and yawning. When prolonged eye closure or yawning is detected, the system issues an alert to the driver. It leverages OpenCV and Dlib libraries for implementation. Additionally, the system detects driver distraction by monitoring head swivels, counting left and right movements. It's worth noting that Dlib may yield facial landmark deviations for non-frontal faces. To mitigate this, the study selected 69 videos from the YawDD database, focusing on near-frontal face views, achieving a 91.3% accuracy rate. © 2023 IEEE.

关键词： Vehicle safety

来源：评论

学校读者我要写书评

暂无评论

Applying system Thinking to Learn from Accident of Modern Automatic Control systems

引用

Chinese Journal of Electronics 2025年第2期23卷 409-414页

作者： Ru Niu Yuan Cao Xiaocheng Ge Tao Tang State Key Laboratory of Rail Traffic Control and Safety Beijing Jiaotong University Beijing China School of Electronics and Information Engineering National Engineering Research Center of Rail Transportation Operation and Control System Beijing Jiaotong University Beijing China Department of Computer Science University of York York UK

Modern societies increasingly rely on automatic control systems. These systems are hardly pure technical systems; instead they are complex socio-technical systems, which consist of technical elements and social components. It is necessary to have a systematic approach to analyze these systems because it is growing evidence that accidents from these systems usually have complex causal factors which form an interconnected network of events, rather than a simple cause-effect chain. We take railway Train control systems (TCS) as an example to demonstrate the importance of the socio-technical approach to analyze the system. The paper presents an investigation of recent high-speed railway accident by applying STAMP - one of the most notable socio-technical system analysis techniques, outlines improvements to the system which could avoid similar accidents in the future. We also provide our valuable feedback for the use of STAMP.

关键词： Sociotechnical systems Analytical models systematics Railway accidents Transportation Medical services Rail transportation systems thinking Research and development Accidents

来源：评论

学校读者我要写书评

暂无评论

Quality-Oriented Task Assignment for Heterogeneous Users in Mobile Crowdsensing

Quality-Oriented Task Assignment for Heterogeneous Users in ...

引用

International Conference on computer Supported Cooperative Work in Design

作者： Kang Chen Peng Li Weiyi Huang Lei Nie Haizhou Bao Qin Liu School of Computer Science and Technology Wuhan University of Science and Technology China Hubei Province Key Laboratory of Intelligent Information Processing and Real-Time Industrial System China School of Cyber Science and Engineering Wuhan University China

Mobile crowdsensing (MCS) is a potential technology for large-scale data collection. This technology requires the platform to recruit users to complete tasks in specific areas. A vital issue in MCS is task assignment, and most existing task assignment efforts consider only a single user type, which is not reasonable in real scenarios. Task assignment becomes more complicated when the platform tries to assign tasks to heterogeneous users with a limited budget in the platform. In this paper, we consider a quality-oriented task assignment for heterogeneous users problem. Professional users have a high sensing quality with a high cost, and normal users have a low sensing quality with a low cost First, we model the sensing capabilities and the costs of different users and formulate the quality-oriented task assignment for heterogeneous users problem. Then, we design the integer linear programming form and prove that the problem is NP-hard. By verifying the submodularity of the objective function, we present a greedy algorithm. Considering the inefficiency of the algorithm, we design two genetic algorithms to improve the total sensing quality. Finally, we evaluate the proposed algorithms under different cost cases based on a real dataset The results show that our proposed algorithm performs well under different cost distribution scenarios.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Two-stage Vehicle Pair Dispatch in Multi-hop Ridesharing

Two-stage Vehicle Pair Dispatch in Multi-hop Ridesharing

引用

International Conference on computer Supported Cooperative Work in Design

作者： Xiaobo Wei Peng Li Weiyi Huang Zhiyuan Liu Qin Liu School of Computer Science and Technology Wuhan University of Science and Technology China Hubei Province Key Laboratory of Intelligent Information Processing and Real-Time Industrial System China School of Cyber Science and Engineering Wuhan University China

Ridesharing benefits the economy and the environment. In multi-hop ridesharing, passengers are permitted to switch vehicles within a single trip, extending the flexibility of conventional ridesharing. Nonetheless, vehicle dispatch is a difficult issue in multi-hop ridesharing. We subdivide the vehicle dispatching problem into the vehicle pairing problem and the request selection problem within a vehicle pair. To address these subproblems, we propose a two-stage framework for vehicle pair dispatching. In the initial stage, we model the vehicle pairing problem as a maximum vehicle-vehicle matching problem in a general graph, which differs from the conventional vehicle-request matching problem in a bipartite graph. The vehicle pairing algorithm is proposed to efficiently solve the vehicle pairing problem. In the second stage, we model the request selection problem as a multidimensional knapsack problem (d-KP) and propose an LP-relaxation request selection algorithm with an approximation ratio 1/5. Experiments conducted on a real-world dataset demonstrate the economic benefit of our proposed two-stage framework.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Contactless Respiratory Rate Monitoring Using FMCW Radar

Contactless Respiratory Rate Monitoring Using FMCW Radar

引用

2022 International Symposium on Electronics and Smart Devices, ISESD 2022

作者： Widiyasari, Diyah Ahmadi, Nur Adiono, Trio Purwarianti, Ayu Purnomo, Ariana Tulus Lin, Ding-Bing Bandung Institute of Technology School of Electrical Engineering and Informatics Bandung40132 Indonesia Bandung40132 Indonesia National Taiwan University of Science and Technology Department of Electronic and Computer Engineering Taipei Taiwan Sampoerna University Departement of Information System Jakarta Selatan12780 Indonesia

ISBN: (数字)9798350396607

ISBN: (纸本)9798350396607

Vital sign monitoring plays a very important role in the early diagnosis and prognosis of a disease, as well as for patient health management. Among other vital signs, respiratory rate (RR) is rarely recorded and often forgotten despite the fact that it is an essential predictor of serious events such as cardiac arrest and deteriorating clinical condition. In this paper, we propose a frequency-modulated continuous-wave (FMCW) radar for contactless RR monitoring. We employ FFT to calculate RR in frequency-domain and combine it with Kalman filter to smooth out the estimated RR. We use a public dataset measuring RR from 50 children under 13 years of age recorded with 60 GHz FMCW radar. The experimental results show that the proposed method yields MAE and RMSE of 3.20 and 4.10, respectively, with respect to the reference device. Incorporating Kalman filter reduces the percentage of error by 14.83% and 13.36% compared to without it. The overall results suggest that radar-based technology can be potentially used for contactless RR monitoring which is suitable for those who cannot use contact-based monitoring devices. © 2022 IEEE.

关键词： Kalman filters

来源：评论

学校读者我要写书评

暂无评论

On Privacy-Preserving Task Assignment for Heterogeneous Users in Mobile Crowdsensing

On Privacy-Preserving Task Assignment for Heterogeneous User...

引用

International Conference on computer Supported Cooperative Work in Design

作者： Ji Zhang Peng Li Weiyi Huang Lei Nie Haizhou Bao Qin Liu School of Computer Science and Technology Wuhan University of Science and Technology China Hubei Province Key Laboratory of Intelligent Information Processing and Real-Time Industrial System China School of Cyber Science and Engineering Wuhan University China

Task assignment is a key challenge in mobile crowd-sensing because of the varying capabilities of crowd users. Location-based task assignment schemes require users to upload their location to an untrusted platform, which raises many privacy concerns. However, stronger privacy preservation may lead to lower system utility. It is challenging to maximize system utility under privacy preservation for users. In this paper, we propose a privacy-preserving task assignment for heterogeneous users (PTAH) problem in mobile crowdsensing. Specifically, we divide users into two groups: private users with location privacy requirements and public users without location privacy requirements. We first design a privacy-preserving mechanism to obfuscate the actual location of private users. Then we construct a relationship graph based on the locations between users and tasks. We prove that the PTAH problem is an NP-hard problem, so to maximize the system utility, we propose an approximation algorithm based on the greedy algorithm. Then we propose a multi-thread cooperative simulated annealing algorithm to search for a better approximate solution. Finally, we conducted simulations based on the widely-used real-world Roma dataset. The results show that our proposed algorithm consistently outperforms other baseline algorithms.

关键词：

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：