检索结果-内蒙古大学图书馆

Exploring the Robustness: Hierarchical Federated Learning Framework for Object Detection of UAV Cluster

IEEE Transactions on Mobile Computing 2025年

作者： Li, Xingyu Zhang, Wenzhe Liu, Linfeng Xu, Jia Nanjing University of Posts and Telecommunications School of Computer Science and Technology Nanjing210023 China Jiangsu Key Laboratory of Big Data Security and Intelligent Processing Nanjing210023 China

The deployment of Unmanned Aerial Vehicle (UAV) cluster is an available solution for object detection missions. In the harsh environment, UAV cluster could suffer from some significant threats (e.g., forest fire hazards, electromagnetic interference, and ground-to-air attacks), which could lead to the destruction of UAVs and loss of data. To this end, we propose a Hierarchical Federated Learning Framework for Object Detection (HFL-OD) to enhance the robustness of UAV cluster conducting object detection missions. In HFL-OD, UAVs are grouped through a Three-Dimensional (3D) graph coloring method, and an intragroup backup mechanism is provided to prevent the data loss caused by the destruction of UAVs. Besides, a dynamic server selection mechanism deals with the potential destruction of servers (cluster server and group servers) by adaptively reassigning the server roles. To further improve the robustness and mission efficiency of UAV cluster, a twotier federated learning framework is introduced to make a proper trade-off between object detection accuracy and communication/computational overhead. This framework is built on the concept of hierarchical federated learning by implementing both intragroup parameter aggregation and global parameter aggregation. Extensive simulations and comparisons demonstrate the superior performance of our proposed HFL-OD, i.e., the robustness of UAV cluster conducting object detection missions can be significantly improved, and the communication/computational overhead is effectively reduced. © 2002-2012 IEEE.

关键词： Fire hazards

来源：评论

学校读者我要写书评

暂无评论

Multi-source Infection Pattern Mining Algorithms over Moving Objects 25

Multi-source Infection Pattern Mining Algorithms over Moving...

引用

25th IEEE International Conference on computer Supported Cooperative Work in Design, CSCWD 2022

作者： Chen, Yu Dai, Hua Yang, Geng Chen, Yanli Nanjing University of Posts and Telecommunications School of Computer Science Nanjing210003 China Jiangsu Key Laboratory of Big Data Security and Intelligent Processing Nanjing210003 China

ISBN: (纸本)9781665405270

Using the trajectory data of moving objects to analyze and study the infection mode of viruses or germs has practical application value. The definition of infection pattern in existing works only considers one-to-one infection mode rather than many-to-one mode, and thus some infection events could be ignored. This paper presents multi-source infection pattern mining algorithms oriented to moving objects for the first time. The multi-source infection event (MSIE) is defined. On the basis of MSIE, the multi-source infection pattern mining algorithm (MIPM) is proposed, which uses sliding window mechanism. The sliding window is used to record the set of candidate infection source that may infect each normal object at each time point, and then the determined infection events are mined. To improve performance, an optimized mining algorithm (MIPM+) based on the R-tree index is proposed, which can reduce the number of objects to be detected at each time point. The experimental results show that our proposed multi-source infection pattern mining algorithms can mine more potential infection events. © 2022 IEEE.

关键词： Decision trees

来源：评论

学校读者我要写书评

暂无评论

Learning Disentangled Latent Factors for Individual Treatment Effect Estimation Using Variational Generative Adversarial Nets 25

Learning Disentangled Latent Factors for Individual Treatmen...

引用

25th IEEE International Conference on computer Supported Cooperative Work in Design, CSCWD 2022

作者： Bao, Qingsen Mao, Zeyong Chen, Lei Nanjing University of Posts and Telecommunications School of Computer Science Nanjing China Nanjing University of Posts and Telecommunications Jiangsu Key Laboratory of Big Data Security and Intelligent Processing Nanjing China

ISBN: (纸本)9781665405270

Estimating individual treatment effect (ITE) is a challenging task due to the need for individual potential outcomes to be learned from biased data and counterfactuals are inherently unobservable. Some researchers propose to use generative adversarial approaches to infer the counterfactual outcomes based on the distribution of factual outcomes. However, these methods assume that complete confounding factors are observed, and simply treat all observed variables as confounding factors, ignoring identification of possible instrumental factors and adjustment factors, which will bring large deviation to ITE estimation when facing biased data. To address these issues, we propose a novel Variational Generative Adversarial Nets for ITE estimation by designing the collaborative learning strategy with Variational AutoEncoder (VAE) and Generative Adversarial Nets (GAN). Specifically, we employ VAE to infer the latent representations of observed variables to access complete latent factors while using GAN to infer unseen counterfactual outcomes and guide VAE for disentangling these latent factors into three sets corresponding to the instrumental, confounding, and adjustment factors. Then the disentangled latent confounding factors can be used to further control data bias using an adaptive weighting scheme. Extensive experiments on real and synthetic data demonstrate learning disentangled latent factors for ITE estimation is effective, and our method has excellent performance even with high data bias. © 2022 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Privacy Protection for Image Sharing Using Reversible Adversarial Examples

Privacy Protection for Image Sharing Using Reversible Advers...

引用

IEEE International Conference on Communications (ICC)

作者： Ping Yin Wei Chen Jiaxi Zheng Yiting Zhang Lifa Wu School of Computer Science Nanjing University of Posts and Telecommunications Nanjing China Jiangsu Key Laboratory of Big Data Security and Intelligent Processing Nanjing China

ISBN: (数字)9781728190549

ISBN: (纸本)9781728190556

Online image sharing on social media platforms faces information leakage due to deep learning-aided privacy attacks. To avoid these attacks, this paper proposes a privacy protection mechanism for image sharing without changing the visual effect, which is based on reversible adversarial examples. Specifically, social media platform users can change the class activation feature to convert the original image into an adversarial image before sharing. When users want to restore the adversarial image to the original image, they can use an improved generative adversarial network model to restore it. The experimental results prove that the conversion model in this paper can effectively prevent privacy attacks from analyzing and stealing users' private information while having no visual impact. At the same time, the proposed restoration model can restore the adversarial examples with high accuracy.

关键词： Training Privacy Image recognition Social networking (online) Perturbation methods Visual effects Information leakage

来源：评论

学校读者我要写书评

暂无评论

Robust Federated Learning for Privacy Preservation and Efficiency in Edge Computing

引用

IEEE Transactions on Services Computing 2025年

作者： Zhou, Hao Dai, Hua Yang, Geng Xiang, Yang Nanjing University of Posts and Telecommunications School of Computer Science Jiangsu Security and Intelligent Processing Lab of Big Data Nanjing210023 China Swinburne University of Technology Department of Computer Science HawthornVIC3122 Australia

Federated Learning (FL) has emerged as a key enabler of privacy-preserving distributed model training in edge computing environments, crucial for service-oriented applications such as personalized healthcare, smart cities, and intelligent assistants. However, existing privacy-preserving FL methods are susceptible to multiple privacy leakage attacks (MPLA), where adversaries infer sensitive information through repeated gradient updates. This paper proposes a Robust and Communication-Efficient Federated Learning (RCFL) framework designed to enhance privacy protection and communication efficiency in edge-based service environments. RCFL integrates a global privacy-preserving mechanism with an innovative privacy encoding strategy that minimizes privacy risks over multiple data releases while significantly reducing communication overhead. The proposed framework's theoretical analysis demonstrates its ability to maintain differential privacy across numerous interactions, ensuring robust model convergence and efficiency. Experimental results using MNIST and CIFAR-10 datasets reveal that RCFL can lower the MPLA success rate from 88.56% to 42.57% compared to state-of-the-art methods, while reducing communication costs by over 90%. These findings underscore RCFL's potential to enhance security, efficiency, and scalability in service-oriented edge computing applications. © 2008-2012 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Decomposing Source Codes by Program Slicing for Bug Localization

Decomposing Source Codes by Program Slicing for Bug Localiza...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Jian Yong Ziye Zhu Yun Li Jiangsu Key Laboratory of Big Data Security and Intelligent Processing School of Computer Science Nanjing University of Posts and Telecommunications

Bug localization, which aims to automatically locate buggy source code files based on the given bug report, is a critical yet time-consuming task in the software engineering field. Existing advanced bug localization methods have successfully leveraged deep learning to bridge the lexical gap between bug reports and source code files at the semantic level. These methods usually first build the entire source code file semantic representation and then match it with the bug report. However, the bug described in the bug report may be related to only part of the source code file semantics. Directly constructing a semantic representation of the entire source code file would increase the difficulty of semantic matching between bug reports and source code files. In this paper, we propose a novel model named S-BugLocator, which decomposes source code file with the help of program slicing. Especially, our proposed S-BugLocator incorporates two distinctly structured slice feature extraction components in processing source code files to cope with the significant discrepancy between multi-row slices and single-row slices. For each multi-row slice, a CNN and Bi-LSTM network is firstly employed to extract its semantics and then a keywords supervised attention mechanism is designed to build its semantic representation by focusing on slices that have strong relevance with the bug report. For each single-row slice, the semantic representation is obtained by fusing word embeddings in single-row slices. The experimental results on four real-world large-scale projects indicate that our proposed model outperforms existing state-of-the-art bug localization methods.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Community Detection Based on Deep Network Embedding with Dual Self-supervised Training 3rd

Community Detection Based on Deep Network Embedding with Dua...

引用

3rd International Conference on big data and security, ICBDS 2021

作者： Chen, Yunfang Mao, Haotian Wang, Li Zhang, Wei School of Computer Science Nanjing University of Posts and Telecommunications Jiangsu Nanjing210023 China Jiangsu Key Laboratory of Big Data Security and Intelligent Processing Nanjing University of Posts and Telecommunications Jiangsu Nanjing210023 China

ISBN: (纸本)9789811908514

We propose a community discovery method based on deep auto-encoding (DGAE_DST). Firstly, we use the pre-trained two-layer neural network and k-means algorithm to initialize the centroid vector, and then use the DNN module and the GCN module to capture the structure features and attributes respectively. The attribute information is fed to the GCN layer to integrate the structure information and attributes. Finally, DNN and GCN respectively output the reconstructed adjacency matrix and feature matrix, and minimize their loss from the original matrix. The experimental results on four datasets show that the DGAE_DST method has higher stability and accuracy. © 2022, Springer Nature Singapore Pte Ltd.

关键词： Complex networks

来源：评论

学校读者我要写书评

暂无评论

Design Guidance for Lightweight Object Detection Models 3rd

Design Guidance for Lightweight Object Detection Models

引用

3rd International Conference on big data and security, ICBDS 2021

作者： Wang, Rui Wang, Xueli Chen, Yunfang Zhang, Wei School of Computer Science Nanjing University of Posts and Telecommunications Jiangsu Nanjing210023 China Jiangsu Key Laboratory of Big Data Security and Intelligent Processing Nanjing University of Posts and Telecommunications Jiangsu Nanjing210023 China

ISBN: (纸本)9789811908514

The lightweight target detection model is deployed in an environment with limited computing power and power consumption, which is widely used in many fields. Most of the current lightweight technologies only focus on a few steps of the model implementation and lack a global perspective. Therefore, this paper proposes a general lightweight model implementation framework, including network construction indicators, lightweight backbone network design, and model optimization. By analyzing the complexity indicator of network structure, the factors that affect network performance such as depth and width are summarized. On this basis, combine the One-Shot Aggregation (OSA) idea and Cross-Stage Partial Network (CSPNet) transformation to construct a general lightweight detection network CSPOSA. Further specific optimization strategies are proposed to prune the network structure and training process. For the network structure, the width and depth of the network are adjusted, the amount of parameters of the model is compressed. The training process is divided into the first, middle and last three stages to improve the detection performance of the model without adding extra computation. Taking embedded platform helmet detection as the experimental scene, the parameter amount of the realized model is 1/10 of the mainstream models YOLOv3 and YOLOv4, and the detection accuracy is similar, so it is more suitable for deployment on devices with limited computing power. © 2022, Springer Nature Singapore Pte Ltd.

关键词： Computing power

来源：评论

学校读者我要写书评

暂无评论

PM$^{2}$2VE: Power Metering Model for Virtualization Environments in Cloud data Centers

引用

IEEE Transactions on Cloud Computing 2023年第3期11卷 3126-3138页

作者： Ziyu Shen Xusheng Zhang Zheng Liu Yun Li School of Computer Science Nanjing University of Posts and Telecommunications Nanjing China Jiangsu Key Laboratory for Big Data Security and Intelligent Processing School of Computer Science Nanjing University of Posts and Telecommunications Nanjing China

Virtualization technologies provide solutions for cloud computing. Virtual resource scheduling is a crucial task in data centers, and the power consumption of virtual resources is a critical foundation of virtualization scheduling. Containers are the smallest unit of virtual resource scheduling and migration. Although many practical models for estimating the power consumption of virtual machines (VMs) have been proposed, few power estimation models of containers have been put forth. In this article, we propose a fast-training piecewise regression model based on a decision tree for VM power metering and estimate the power of containers configured on the VM by treating the container as a group of processes on the VM. We select appropriate features from the collected metrics of VMs/containers to help our model fit the nonlinear relationship between power and features well. Besides, we optimize the leaf nodes of the regression tree, realizing the effective power metering of virtualization environments. We evaluate the proposed model on 13 tasks in PARSEC and compare it with several commonly used models in data centers. The experimental results prove the effectiveness of the proposed model, and the estimated power of containers is in line with expectations.

关键词： Containers data centers Power demand Estimation data models Virtualization Servers

来源：评论

学校读者我要写书评

暂无评论

DBSSL: A Scheme to Detect Backdoor Attacks in Self-Supervised Learning Models

引用

IEEE Transactions on Dependable and Secure Computing 2025年

作者： Huang, Yuxian Yang, Geng Yuan, Dong Yu, Shui Nanjing University of Posts and Telecommunication College of Computer Science and Software China Jiangsu Key Laboratory of Big Data Security and Intelligent Processing China University of Sydney School of Electrical and Information Engineering Australia University of Technology Sydney School of Computer Science Australia

Recently, self-supervised learning has garnered significant attention for its ability to extract high-quality features from unlabeled data. However, existing research indicates that backdoor attacks can pose significant threats to self-supervised learning. Additionally, due to the substantial training overhead, traditional backdoor detection methods in supervised learning, such as meta-learning, are not well-suited for self-supervised learning. Current backdoor detection methods for self-supervised learning can only defend against backdoor attacks triggered by patch. To address this challenge, we proposes DBSSL, a scheme that can efficiently detect backdoor attacks in self-supervised learning models. In DBSSL, we connect backdoor attacks with adversarial attacks and use adversarial perturbations for detection. Specifically, we first conduct targeted adversarial attacks on the samples, classifying them into different classes. Then, we employ the Z-Score method to analyze these adversarial perturbations. If outliers are present, we can infer that the self-supervised learning model has been backdoored. Unlike previous works, our scheme is better suited for downstream users with constrained computation resources and exhibits excellent detection capabilities for backdoor attacks caused by patch -type or global triggers. Abundant theoretical analysis has demonstrated the feasibility of our scheme. Moreover, extensive experiments show that our method performs well in detecting prevalent backdoor attacks in self-supervised learning, achieving detection accuracy exceeding 95%. © 2004-2012 IEEE.

关键词： Self-supervised learning

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：