检索结果-内蒙古大学图书馆

IEEE International Conference on software Reliability engineering Workshops (ISSRE Wksp)

作者： Qingran Su Xingze Li Yuming Ren Xulang Ouyang Chunming Hu Yongfeng Yin School of Computer Science and Engineering Beihang University Beijing China School of Software Beihang University Beijing China

ISBN: (数字)9798350367041

ISBN: (纸本)9798350367058

Safety testing is a key method to ensure software quality. But the quality of the test depends on the level of the test engineers. The method of generating safety test cases based on state diagrams often perform poorly due to lack of personnel experience. Meanwhile, the manual construction of test cases constrains the improvement of safety testing efficiency. This article utilizes LLMs to achieve automated extension and testing of safety state diagrams. The main contributions include extracting safety properties from aviation standards and extending the basic state diagram with LLMs to enable safety testing capabilities. In addition, a genetic algorithm combining edge first depth traversal and LLMs optimization was proposed to improve the efficiency of test case generation. Compared to classical genetic algorithms, the average execution round achieved a 90% reduction. Our LLMs optimization strategy can also enhance other improved genetic algorithms, further improving the efficiency of safety testing.

关键词： Large language models software quality Manuals Safety software reliability Personnel Optimization Standards Testing Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

Physical Layer Encrypted Maritime Communications Utilizing UAV-Enabled Virtual Antenna Array

Physical Layer Encrypted Maritime Communications Utilizing U...

引用

IEEE International Conference on Communications (ICC)

作者： Jiawei Huang Aimin Wang Geng Sun Jiahui Li Xiaoya Zheng College of Software Engineering and College of Computer Science and Technology Jilin University Changchun China School of Computer Science and Engineering Nanyang Technological University Singapore Singapore

ISBN: (数字)9781728190549

ISBN: (纸本)9781728190556

Maritime wireless communications as the promising applications gradually spur people's interest. However, they are suffering from critical security issues due to the feature of open channels. This paper propose to utilize a group of unmanned aerial vehicles (UAVs) as a jammer to achieve physical layer security during the maritime wireless communications. Specifically, these UAVs form a maritime UAV-enabled virtual antenna array (UVAA), which can be allowed to transmit jamming signals to the eavesdropper. In the designed system, we formulate a maritime wireless communication multi-objective optimization problem (MWCMOP) to simultaneously achieve the maximization of the signal-to-interference-plus-noise ratio (SINR) of the legal vessel, minimization of the SINR of the eavesdropping vessel, and minimization of the total flying energy cost of UAVs. It is challenging to solve the formulated MWCMOP since it is complex and NP-hard. Thus, we present a novel improved evolutionary algorithm to deal with the problem. Simulation results show that the presented algorithm is superior to other peer algorithms and it is capable of searching a transmission-maximizing flight scheme.

关键词： Wireless communication Simulation Interference Autonomous aerial vehicles Cryptography Jamming Maritime communications

来源：评论

学校读者我要写书评

暂无评论

Artifact-Focused Attention Networks for Robust Detection of High-Quality Deepfake Videos in Short-Frame Scenarios

Artifact-Focused Attention Networks for Robust Detection of ...

引用

Information Systems and Computing Technology (ISCTech), International Conference on

作者： Yiming Liu Xiaoru Wang School of Computer Science(National Pilot Software Engineering School) Beijing University of Posts and Telecommunications Beijing China

ISBN: (数字)9798350379860

ISBN: (纸本)9798350379877

Deepfake technology, driven by advances in generative adversarial networks (GANs) and diffusion models, has achieved unprecedented realism, posing significant challenges to video authentication. This paper presents an Artifact-Focused Attention Network (AFAN) designed to detect high-quality deepfake videos, particularly in short-frame scenarios with limited spatial information. By leveraging explicit attention mechanisms and artifactenhancement strategies, the proposed framework effectively identifies subtle spatial inconsistencies and local artifacts introduced by generative models. Experimental results on FF++ and other datasets demonstrate that AFAN achieves state-of-the-art performance, surpassing existing methods in accuracy and robustness. This research provides a practical solution to the growing threat of deepfake videos in real-world applications.

关键词： Deepfakes Attention mechanisms Accuracy Scalability Perturbation methods Generative adversarial networks Robustness Real-time systems Resilience Information systems

来源：评论

学校读者我要写书评

暂无评论

Disease Diagnosis Based on Multi-View Contrastive Learning for Electronic Medical Records

引用

IAENG International Journal of Applied Mathematics 2023年第3期53卷

作者： Zhang, Zhengkang Yang, Dan Zhang, Yu School of Computer Science and Software Engineering University of Science and Technology Liaoning Anshan China School of computer Science and Software Engineering University of Science and Technology Liaoning Anshan China China Telecom Digital Intelligence Technology Co. Ltd Beijing China

Disease diagnosis based on electronic medical records (EMRs) is one of the important research contents of intelligent healthcare. In recent years, disease diagnosis based on heterogeneous graph neural networks has received increasing attention. However, disease diagnosis tasks often suffer from the lack of labeling information due to the high cost of manual labeling. And existing disease diagnosis models based on heterogeneous graph neural networks ignore the correlation between different meta-paths. Therefore, we propose a disease diagnosis model based on multi-view contrastive learning (MVCDD). MVCDD uses medical data from electronic medical records to construct medical heterogeneous graphs, and uses a fixed-depth random walk method to obtain semantic subgraphs defined by multiple meta-paths. Meanwhile, we introduce the inter-view contrastive learning task to model the correlation between different meta-paths. MVCDD optimizes the patient representations by combining intra-view and inter-view contrastive learning tasks jointly. Extensive experiments are conducted on the MIMIC-III dataset. The experimental results on the MIMIC-III dataset demonstrate that MVCDD outperforms other baselines and effectively improves the performance of disease diagnosis. © (2023), (International Association of Engineers). All Rights Reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Denoising Diffusion Model-Based Channel Estimation in IRS-Assisted ISAC Systems 16

Denoising Diffusion Model-Based Channel Estimation in IRS-As...

引用

16th International Conference on Wireless Communications and Signal Processing, WCSP 2024

作者： Yi, Lan Miao, Jiansong Zhang, Tao Yuan, Zhenhui Kang, Jiawen Niyato, Dusit Beijing University of Posts and Telecommunications Beijing Laboratory of Advanced Information Networks Beijing China School of Software Engineering Beijing Jiaotong University Beijing China School of Engineering University of Warwick CV4 7AL United Kingdom School of Automation Guangdong University of Technology Guangzhou China School of Computer Science and Engineering Nanyang Technological University Singapore Singapore

ISBN: (纸本)9798350390643

In Integrated Sensing and Communications (ISAC) systems assisted by Intelligent Reflecting Surfaces (IRS), precise channel estimation is crucial for optimal system performance and maximizing the potential benefits. The estimation problem, however, becomes intricate due to complex channel coupling and specific environmental conditions. This paper addresses these challenges by formulating channel estimation as a denoising problem and proposing a novel diffusion-based framework. This approach mitigates the impact of ill-posedness, which is typically problematic for general neural network (NN)-based techniques. By modeling the joint distribution and sampling from the learned posterior distribution, we accurately recover the channel coefficients from noisy pilot-based observations. Simulation results demonstrate significant performance improvements of the proposed method compared to both least-squares and NN-based benchmarks. Moreover, thanks to the carefully designed architecture and residual prediction strategy, the number of sampling steps in the diffusion reverse process is greatly reduced, enabling real-time CSI acquisition. © 2024 IEEE.

关键词： Least squares approximations

来源：评论

学校读者我要写书评

暂无评论

Lithological Classification Using Densely Connected Convolution Network on Landsat-9 and Aster Datasets in a Semi-Arid Environment 20

Lithological Classification Using Densely Connected Convolut...

引用

20th International computer Conference on Wavelet Active Media Technology and Information Processing, ICCWAMTIP 2023

作者： Appiah-Twum, Michael Wenbo, Xu Mawuli, Cobbinah B. Atandoh, Peter School of Resources and Environment University of Electronic Science and Technology of China Sichuan Chengdu611731 China University of Electronic Science and Technology of China Huzhou313001 China School of Computer Science and Engineering University of Electronic Science and Technology of China Sichuan Chengdu611731 China School of Information and Software Engineering University of Electronic Science and Technology of China Sichuan Chengdu611731 China

ISBN: (纸本)9798350318982

Remote sensing-based geological interpretation provides insight from numerous sensors. That notwithstanding, high-dimensional feature learning requires full-scale diagnosis of intrinsic geological features in remotely sensed datasets. This study proposes a multi-sensor fusion approach. Integrating Landsat-9 and ASTER spectral features with DenseNet for lithological classification, displaces RF and KNN by 2.12 % and surpasses ResNet50 and Xception with a classification accuracy of 83.89%. Furthermore, this study underscores the effectiveness of merging new and old generation sensors for geological interpretation. © 2023 IEEE.

关键词： Landsat

来源：评论

学校读者我要写书评

暂无评论

A Deep Reinforcement Learning Perspective on Adaptive Federated Dropout 20

A Deep Reinforcement Learning Perspective on Adaptive Federa...

引用

20th IEEE International Conference on Mobile Ad Hoc and Smart Systems, MASS 2023

作者： Sun, Tonghui Huang, Yan Xie, Zhenzhen Pang, Junjie Chen, Xingyun Cai, Zhipeng Qingdao University College of Computer Science and Technology Qingdao266000 China Kennesaw State University College of Computing and Software Engineering AtlantaGA United States Shandong University School of Computer Science and Technology Qingdao266237 China Georgia State University Department of Computer Science AtlantaGA United States

ISBN: (纸本)9798350324334

Federated learning is a distributed learning solution that achieves high-quality machine learning models while ensuring privacy and collaboration among various end devices. However, different kinds of end devices can lead to an unstable training process due to limited and dynamic communication and computation resources. Federated Dropout is an important technique for mitigating such resource bottlenecks by randomly removing a fixed percentage of activations of model components. However, most of the existing FL dropout methods cannot effectively utilize the distinguished characteristics of different clients or fail to automatically adapt to end device heterogeneity. In this paper, we thus propose Adaptive Federated Dropout with Reinforcement Learning (AFD-RL). AFD-RL employs reinforcement learning technology to adapt to the clients' unbalanced computation and communication resources and provide a personalized dropout strategy for each client. By automatically selecting suitable model components for each client in each communication round, AFD-RL achieves better accuracy with fewer communication rounds. Extensive experiments validate the effectiveness of the proposed approach in improving training efficiency and inference accuracy. © 2023 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

DeepDir: a deep learning approach for API directive detection

引用

science China(Information sciences) 2021年第9期64卷 238-240页

作者： Jingxuan ZHANG He JIANG Shuai LU Ge LI Xin CHEN College of Computer Science and Technology Nanjing University of Aeronautics and Astronautics State Key Laboratory for Novel Software Technology Nanjing University Key Laboratory of Complex Systems Modeling and Simulation Ministry of Education School of Software Dalian University of Technology School of Electronic Engineering and Computer Science Peking University School of Computer Science and Technology Hangzhou Dianzi University

Dear editor,software developers tend to reuse existing libraries to facilitate their development process and implement certain functionalities by invoking application programming interfaces(APIs) [1]. However, it remains a challenging task for developers to correctly use APIs [2], so they often consult API learning resources [3, 4]. As one of the most important API learning resources,

关键词： API Specification API Directive Deep Learning Text Classification Imbalanced Learning

来源：评论

学校读者我要写书评

暂无评论

Task assignment for social-oriented crowdsourcing

引用

Frontiers of computer science 2021年第2期15卷 39-49页

作者： Gang WU Zhiyong CHEN Jia LIU Donghong HAN Baiyou QIAO School of Computer Science and Engineering Northeastern UniversityShenyang110004China State Key Laboratory for Novel Software Technology Nanjing UniversityNanjing210023China

Crowdsourcing has become an efficient measure to solve machine-hard problems by embracing group wisdom,in which tasks are disseminated and assigned to a group of workers in the way of open *** social relationships formed during this process may in turn contribute to the completion of future *** this sense,it is necessary to take social factors into consideration in the research of ***,there is little work on the interactions between social relationships and crowdsourcing *** this paper,we propose to study such interactions in those social-oriented crowdsourcing systems from the perspective of task assignment.A prototype system is built to help users publish,assign,accept,and accomplish location-based crowdsourcing tasks as well as promoting the development and utilization of social relationships during the ***,in order to exploit the potential relationships between crowdsourcing workers and tasks,we propose a“worker-task”accuracy estimation algorithm based on a graph model that joints the factorized matrixes of both the user social networks and the history“worker-task”*** the worker-task accuracy estimation matrix,a group of optimal worker candidates is efficiently chosen for a task,and a greedy task assignment algorithm is proposed to further the matching of worker-task pairs among multiple crowdsourcing tasks so as to maximize the overall *** with the similarity based task assignment algorithm,experimental results show that the average recommendation success rate increased by 3.67%;the average task completion rate increased by 6.17%;the number of new friends added per week increased from 7.4 to 10.5;and the average task acceptance time decreased by 8.5 seconds.

关键词： crowdsourcing social networks task assignment

来源：评论

学校读者我要写书评

暂无评论

DPRec: Social Recommendation Based on Dynamic User Preferences

IAENG International Journal of Computer Science

引用

IAENG International Journal of computer science 2023年第3期50卷

作者： Hu, Haibo Yang, Dan Zhang, Yu School of Computer Science and Software Engineering University of Science and Technology Liaoning Anshan China School of computer Science and Software Engineering University of Science and Technology Liaoning Anshan China China Telecom Digital Intelligence Technology Co. Ltd Beijing China

Social recommendation learns users' preferences by integrating social information and interaction information to complete the recommendation task. In recent years, social recommendation has begun to model high-order neighbors of users to learn their preferences. However, existing social recommendation models rarely model dynamic user preferences, ignoring the similarity relationships between items and the inconsistency problem of neighbors. Therefore, we propose a social recommendation algorithm based on dynamic user preferences (DPRec), which models dynamic user preferences through a RNN. we use the similarity coefficient to calculate similarity relationships between items and constructs an item similarity network graph. At the same time, DPRec dynamically samples neighbor information of nodes (users and items) to reduce the impact of social inconsistency neighbors on recommendation performance. Based on the different influences of different neighbors, DPRec uses an attention mechanism to learn the influence factors of different neighbors on the target node. The experiments on two public datasets reveal the importance of dynamic user preferences and verify the effectiveness of the algorithm. © (2023), (International Association of Engineers). All Rights Reserved.

关键词： Graph neural networks

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：