检索结果-内蒙古大学图书馆

3rd World Conference on Intelligent and 3D Technologies, WCI3DT 2024

作者： Su, Jianna Department of Mathematics and Computer Engineering Ordos Institute of Applied Technology Inner Mongolia Erdos017010 China

ISBN: (纸本)9789819791231

This article explores the application of computer vision technology in user interface design, particularly through gesture recognition to enhance human–computer interaction. Gestures offer natural and versatile advantages in the process of human–computer interaction. Vision-based gesture recognition technology has broad application potential in virtual environments, sign language recognition, machine control, and other fields. In the task of gesture recognition in user interface design, the recognition results are affected by some irrelevant factors, such as dynamic backgrounds and close-fitting clothing. The attention mechanism is a good way to address the above problems, as it applies weights to key areas to extract essential features, thereby improving the model's performance. Several commonly used attention mechanisms, such as spatial attention, channel attention, and mixed attention mechanisms, can only apply attention to a single-frame image, focusing on specific regions. However, gestures consist of sequences of multiple-frame images, and the recognition results are collectively influenced by multiple-frame gesture images. The aforementioned attention mechanisms cannot meet the requirements effectively. To address this issue, a joint intra-frame and inter-frame attention module has been investigated. The model is able to prevent the impact of irrelevant elements during the gesture identification process because of this module's ability to apply attention concurrently across many consecutive frames and inside a single frame. In order to further user interface design, this study employs convolutional neural networks to include gesture recognition. These results shed light on how computer vision technology may revolutionize UI design and enhance user experiences. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Robustness of generative AI detection: adversarial attacks on black-box neural text detectors

引用

International Journal of Speech Technology 2024年第4期27卷 861-874页

作者： Fishchuk, Vitalii Braun, Daniel Faculty of Electrical Engineering Mathematics and Computer Science University of Twente Enschede Netherlands Department of High-tech Business and Entrepreneurship University of Twente Enschede Netherlands

The increased quality and human-likeness of AI generated texts has resulted in a rising demand for neural text detectors, i.e. software that is able to detect whether a text was written by a human or generated by an AI. Such tools are often used in contexts where the use of AI is restricted or completely prohibited, e.g. in educational contexts. It is, therefore, important for the effectiveness of such tools that they are robust towards deliberate attempts to hide the fact that a text was generated by an AI. In this article, we investigate a broad range of adversarial attacks in English texts with six different neural text detectors, including commercial and research tools. While the results show that no detector is completely invulnerable to adversarial attacks, the latest generation of commercial detectors proved to be very robust and not significantly influenced by most of the evaluated attack strategies. © The Author(s) 2024.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Early Detection of Cyber-Physical Attacks on Electric Vehicles Fast Charging Stations Using Wavelets and Deep Learning

IEEE Transactions on Industrial Cyber-Physical Systems

引用

IEEE Transactions on Industrial Cyber-Physical Systems 2024年 2卷 220-231页

作者： Abu-Nassar, Ahmad M. Morsi, Walid G. Electrical Computer and Software Engineering Department Faculty of Engineering and Applied Science OshawaONL1G 0C5 Canada

Transportation electrification plays an important role in the operation of the smart grid through the integration of the electric vehicle fast charging stations (EVFCSs), which allows the electric vehicles to provide regulation services to the grid through the vehicle-to-grid (V2G) concept. However, such an integration makes smart grid assets prone to cyber vulnerability threats. In this paper, a cyber-physical attack detection approach is developed to early detect such attacks. The proposed approach combines the continuous wavelet transform (CWT) and the convolution neural network (CNN) to provide an effective detection technique. The proposed detection approach has undergone rigorous testing that considered 420 realistic operational scenarios. Unlike in previous work, the proposed detection approach was found to be effective in automatically learning the salient features from the data as well as identifying the frequency bands that hold such features and using them in the classification process. Furthermore, this work investigated the cyber-attack detection accuracy using different time resolutions of smart meters. The results have shown that the proposed approach effectively detects cyber-physical attacks with an accuracy of 99.76% and a low computational time of 1.8 seconds. © 2024 IEEE.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Trustworthy Actionable Perturbations 41

Trustworthy Actionable Perturbations

引用

41st International Conference on Machine Learning, ICML 2024

作者： Friedbaum, Jesse Adiga, Sudarshan Tandon, Ravi Program in Applied Mathematics University of Arizona TucsonAZ United States Department of Electrical and Computer Engineering University of Arizona TucsonAZ United States

Counterfactuals, or modified inputs that lead to a different outcome, are an important tool for understanding the logic used by machine learning classifiers and how to change an undesirable classification. Even if a counterfactual changes a classifier's decision, however, it may not affect the true underlying class probabilities, i.e. the counterfactual may act like an adversarial attack and "fool" the classifier. We propose a new framework for creating modified inputs that change the true underlying probabilities in a beneficial way which we call Trustworthy Actionable Perturbations (TAP). This includes a novel verification procedure to ensure that TAP change the true class probabilities instead of acting adversarially. Our framework also includes new cost, reward, and goal definitions that are better suited to effectuating change in the real world. We present PAC-learnability results for our verification procedure and theoretically analyze our new method for measuring reward. We also develop a methodology for creating TAP and compare our results to those achieved by previous counterfactual methods. Copyright 2024 by the author(s)

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Privacy-preserving explainable AI: a survey

引用

Science China(Information Sciences) 2025年第1期68卷 23-56页

作者： Thanh Tam NGUYEN Thanh Trung HUYNH Zhao REN Thanh Toan NGUYEN Phi Le NGUYEN Hongzhi YIN Quoc Viet Hung NGUYEN School of Information and Communication Technology Griffith University School of Computer and Communication Sciences Ecole Polytechnique Federale de Lausanne Faculty of Mathematics and Computer Science University of Bremen Faculty of Information Technology HUTECH University Department of Computer Science Hanoi University of Science and Technology School of Electrical Engineering and Computer Science The University of Queensland

As the adoption of explainable AI(XAI) continues to expand, the urgency to address its privacy implications intensifies. Despite a growing corpus of research in AI privacy and explainability, there is little attention on privacy-preserving model explanations. This article presents the first thorough survey about privacy attacks on model explanations and their countermeasures. Our contribution to this field comprises a thorough analysis of research papers with a connected taxonomy that facilitates the categorization of privacy attacks and countermeasures based on the targeted explanations. This work also includes an initial investigation into the causes of privacy leaks. Finally, we discuss unresolved issues and prospective research directions uncovered in our analysis. This survey aims to be a valuable resource for the research community and offers clear insights for those new to this domain. To support ongoing research, we have established an online resource repository, which will be continuously updated with new and relevant findings.

关键词： privacy-preserving explainable AI privacy attacks privacy defences PrivEx PPXAI

来源：评论

学校读者我要写书评

暂无评论

Decentralized Multi-Armed Bandit Can Outperform Classic Upper Confidence Bound: A Homogeneous Case over Strongly Connected Graphs 63

Decentralized Multi-Armed Bandit Can Outperform Classic Uppe...

引用

63rd IEEE Conference on Decision and Control, CDC 2024

作者： Zhu, Jingxuan Liu, Ji Zhejiang Lab China Stony Brook University Department of Applied Mathematics and Statistics United States Stony Brook University Department of Electrical and Computer Engineering United States

ISBN: (纸本)9798350316339

This paper studies a homogeneous decentralized multi-armed bandit problem, in which a network of multiple agents faces the same set of arms, and each agent aims to minimize its own regret. A fully decentralized upper confidence bound (UCB) algorithm is proposed for a multi-agent network whose neighbor relations are described by a directed graph. It is shown that the decentralized algorithm guarantees each agent to achieve a lower logarithmic asymptotic regret compared to the classic UCB algorithm, provided the neighbor graph is strongly connected. The improved asymptotic regret upper bound is reciprocally related to the maximal size of a local neighborhood within the network. The roles of graph connectivity, maximum local degree, and network size are analytically elucidated in the expression of regret. © 2024 IEEE.

关键词： Decentralized systems

来源：评论

学校读者我要写书评

暂无评论

Blockchain-based secure and transparent election and vote counting mechanism using secret sharing scheme

引用

Journal of Ambient Intelligence and Humanized Computing 2023年第10期14卷 14009-14027页

作者： Chouhan, Vikas Arora, Anshul Department of Computer Science and Engineering Indian Institute of Technology Roorkee Roorkee India Department of Applied Mathematics Delhi Technological University Delhi India

Privacy and transparency in vote counting are the most prevalent concerns these days due to the involvement of untrusted authorities in the counting process. As a result, the counting process faces significant privacy, trust, and transparency hurdles. Hence, there is a need for an efficient and trusted mechanism to resolve such problems. Blockchain technology has the potential to bring transparency and trust in several applications. Therefore, in this work, we explore blockchain technology in conjunction with a secure partitioning scheme to promote transparency, trust, and privacy between users and participating authorities in a decentralized platform. This paper presents a chaincode-based implementation of our proposed secure and verifiable vote counting mechanism that enables trust and fairness over a decentralized platform. Multiple authorities participate in the vote counting process in a trusted manner to cooperate and coordinate in a decision process over a decentralized platform. Our research exhibits that blockchain technology can eliminate the trust gaps and increase transparency and fairness in the election and vote counting procedure. We register user votes in the blockchain platform based on the secret sharing mechanism to enable fairness and openness between counting authorities. Each vote is recorded into the distributed ledger to support openness and verifiability in our mechanism. The ledger is accessible to every registered user as per the permissioned blockchain policy. We created many authorities in the blockchain network and deployed multiple smart contracts on the Hyperledger platform to analyze the feasibility of our strategy. The performance results are obtained and reported using the Hyperledger Caliper benchmark tool. The results demonstrate that the proposed chaincode-based solution achieves the highest throughput at 200–400 tps for fetching and removing contracts. We achieve the optimal latency of 18.09 s for the vote distribution contract

关键词： Smart contract

来源：评论

学校读者我要写书评

暂无评论

Spatio-temporal Graph-Based Generation and Detection of Adversarial False Data Injection Evasion Attacks in Smart Grids

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第12期5卷 6601-6616页

作者： Takiddin, Abdulrahman Ismail, Muhammad Atat, Rachad Serpedin, Erchin Florida State University Department of Electrical and Computer Engineering FAMU-FSU College of Engineering TallahasseeFL32310 United States Tennessee Tech University Department of Computer Science CookevilleTN38505 United States Lebanese American University Department of Computer Science and Mathematics Beirut1102-2801 Lebanon Texas A&M University Department of Electrical and Computer Engineering College StationTX77843 United States

Smart power grids are vulnerable to security threats due to their cyber-physical nature. Existing data-driven detectors aim to address simple traditional false data injection attacks (FDIAs). However, adversarial false data injection evasion attacks (FDIEAs) present a more serious threat as adversaries, with different levels of knowledge about the system, inject adversarial samples to circumvent the grid's attack detection system. The robustness of state-of-the-art graph-based detectors has not been investigated against sophisticated FDIEAs. Hence, this article answers three research questions. 1) What is the impact of utilizing spatio-temporal features to craft adversarial samples and how to select attack nodes? 2) How can adversaries generate surrogate spatio-temporal data when they lack knowledge about the system topology? 3) What are the required model characteristics for a robust detection against adversarial FDIEAs? To answer the questions, we examine the robustness of several detectors against five attack cases and conclude the following: 1) Attack generation with full knowledge using spatio-temporal features leads to 5%-26% and 2%-5% higher degradation in detection rate (DR) compared to traditional FDIAs and using temporal features, respectively, whereas centrality analysis-based attack node selection leads to 3%-11% higher degradation in DR compared to a random selection;2) Stochastic geometry-based graph generation to create surrogate adversarial topologies and samples leads to 3%-13% higher degradation in DR compared to traditional FDIAs;and 3) Adopting an unsupervised spatio-temporal graph autoencoder (STGAE)-based detector enhances the DR by 5-53% compared to benchmark detectors against FDIEAs. © 2020 IEEE.

关键词： Cyber attacks

来源：评论

学校读者我要写书评

暂无评论

Detecting Abnormal Communication Patterns in IoT Networks Using Graph Neural Networks 13th

Detecting Abnormal Communication Patterns in IoT Networks U...

引用

13th IAPR-TC-15 International Workshop on Graph-Based Representations in Pattern Recognition, GbRPR 2023

作者： Carletti, Vincenzo Foggia, Pasquale Vento, Mario Department of Information Engineering Electrical Engineering and Applied Mathematics University of Salerno Fisciano Italy

ISBN: (纸本)9783031427947

Nowadays, millions of Internet of Things (IoT) devices communicate over the Internet, thus becoming potential targets for cyberattacks. Due to the limited hardware capabilities of these devices, host-based countermeasures are unlikely to be deployed on them, making network traffic analysis the only reasonable way to detect malicious activities. In this paper, we face the problem of identifying abnormal communications in IoT networks using graph-based anomaly detection methods. Although anomaly detection has already been applied to graph-based data, most existing methods have been used for static graphs, with the aim of detecting anomalous nodes. In our case, the graphs represent snapshots of the network traffic, and change with time. In this paper we compare different graph-based methods, and different graph representations of the network traffic, using two large datasets of real IoT data. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Coordinated Motion Planning: Multi-Agent Path Finding in a Densely Packed, Bounded Domain 35

Coordinated Motion Planning: Multi-Agent Path Finding in a D...

引用

35th International Symposium on Algorithms and Computation, ISAAC 2024

作者： Fekete, Sándor P. Kosfeld, Ramin Kramer, Peter Neutzner, Jonas Rieck, Christian Scheffer, Christian Department of Computer Science TU Braunschweig Germany Department of Electrical Engineering and Computer Science Bochum University of Applied Sciences Germany

ISBN: (纸本)9783959773546

We study Multi-Agent Path Finding for arrangements of labeled agents in the interior of a simply connected domain: Given a unique start and target position for each agent, the goal is to find a sequence of parallel, collision-free agent motions that minimizes the overall time (the makespan) until all agents have reached their respective targets. A natural case is that of a simply connected polygonal domain with axis-parallel boundaries and integer coordinates, i.e., a simple polyomino, which amounts to a simply connected union of lattice unit squares or cells. We focus on the particularly challenging setting of densely packed agents, i.e., one per cell, which strongly restricts the mobility of agents, and requires intricate coordination of motion. We provide a variety of novel results for this problem, including (1) a characterization of polyominoes in which a reconfiguration plan is guaranteed to exist;(2) a characterization of shape parameters that induce worst-case bounds on the makespan;(3) a suite of algorithms to achieve asymptotically worst-case optimal performance with respect to the achievable stretch for cases with severely limited maneuverability. This corresponds to bounding the ratio between obtained makespan and the lower bound provided by the max-min distance between the start and target position of any agent and our shape parameters. Our results extend findings by Demaine et al. [13, 14] who investigated the problem for solid rectangular domains, and in the closely related field of Permutation Routing, as presented by Alpert et al. [6] for convex pieces of grid graphs. © Sándor P. Fekete, Ramin Kosfeld, Peter Kramer, Jonas Neutzner, Christian Rieck, and Christian Scheffer.

关键词： Motion planning

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：