检索结果-内蒙古大学图书馆

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Huang, Yihuan Li, Yuanzhe Ren, Yanzhen Tu, Weiping Yang, Yuhong Key Laboratory of Aerospace Information Security and Trusted Computing Ministry of Education China School of Cyber Science and Engineering Wuhan University China National Engineering Research Center for Multimedia Software School of Computer Science Wuhan University China

ISBN: (纸本)9798350368741

Speaker recognition (SR) systems are particularly vulnerable to adversarial example (AE) attacks. To mitigate these attacks, AE detection systems are typically integrated into SR systems. To overcome the limitations of low detection accuracy, poor generalization, and high latency in existing schemes, this paper proposes FreqSense, an AE detection scheme based on frequency distribution features. FreqSense detects a variety of unknown AE attacks with low latency, and provides interpretability in its detection process. The basic idea of FreqSense is that AE typically introduce carefully designed noise in specific frequency bands that are associated with highly distinctive speaker identities. Therefore, leveraging the distributional variations in these frequency bands can effectively distinguish between AE and benign audio. FreqSense models frequency distribution features by integrating time-frequency transformation technology with a self-attention mechanism and employs a neural network-based classifier to distinguish between AE and benign audio. Experimental results show that FreqSense achieves an overall detection accuracy of 99.2%, surpassing state-of-the-art (SOTA) schemes by 27.2%. When confronting unknown AE attacks, FreqSense achieves a detection accuracy of 98.3% with a latency of just 0.0014 seconds. © 2025 IEEE.

关键词： Adversarial Attack Adversarial Example Detection Speaker Recognition

来源：评论

学校读者我要写书评

暂无评论

First Creating Backgrounds Then Rendering Texts: A New Paradigm for Visual Text Blending 27

First Creating Backgrounds Then Rendering Texts: A New Parad...

引用

27th European Conference on Artificial Intelligence, ECAI 2024

作者： Li, Zhenhang Shu, Yan Zeng, Weichao Yang, Dongbao Zhou, Yu Institute of Information Engineering Chinese Academy of Sciences Beijing China TMCC College of Computer Science Nankai University Tianjin China School of Cyber Security University of Chinese Academy of Sciences Beijing China

ISBN: (纸本)9781643685489

Diffusion models, known for their impressive image generation abilities, have played a pivotal role in the rise of visual text generation. Nevertheless, existing visual text generation methods often focus on generating entire images with text prompts, leading to imprecise control and limited practicality. A more promising direction is visual text blending, which focuses on seamlessly merging texts onto text-free backgrounds. However, existing visual text blending methods often struggle to generate high-fidelity and diverse images due to a shortage of backgrounds for synthesis and limited generalization capabilities. To overcome these challenges, we propose a new visual text blending paradigm including both creating backgrounds and rendering texts. Specifically, a background generator is developed to produce high-fidelity and text-free natural images. Moreover, a text renderer named GlyphOnly is designed for achieving visually plausible text-background integration. GlyphOnly, built on a Stable Diffusion framework, utilizes glyphs and backgrounds as conditions for accurate rendering and consistency control, as well as equipped with an adaptive text block exploration strategy for small-scale text rendering. We also explore several downstream applications based on our method, including scene text dataset synthesis for boosting scene text detectors, as well as text image customization and editing. Code and model will be available at https://***/Zhenhang-Li/GlyphOnly. © 2024 The Authors.

关键词： Rendering (computer graphics)

来源：评论

学校读者我要写书评

暂无评论

Leveraging XGBoost Machine Learning Algorithm for Common Vulnerabilities and Exposures (CVE) Exploitability Classification

Leveraging XGBoost Machine Learning Algorithm for Common Vul...

引用

International Conference on Computation System and Information Technology for Sustainable Solutions (CSITSS)

作者： Meryn Babu Namruth Reddy S Minal Moharir Mohana Computer Science and Engineering (Cyber Security) RV College of Engineering® Bengaluru India Security Engineer NVIDIA Bengaluru India

ISBN: (数字)9798331505462

ISBN: (纸本)9798331505479

The problem of exploitation is highly critical in cybersecurity. Therefore, it is very important for the prioritization of security patching efforts and risk mitigation to accurately predict the exploitability of a given vulnerability using the Common Vulnerabilities and Exposures System. In this paper XGBoost ML model is used for the classification of CVEs. According to the analysis of the data, compared with others, network-based vulnerabilities are still predominant at 85%, while local and adjacent network vectors stood at 12.9% and 1.9%, respectively. The estimate of the most affected operating systems was on Linux and Microsoft, with about 37,500 and 11,000 CVEs, respectively. Most of the vulnerabilities resided in cryptographic issues, cross-site scripting, restriction on memory buffers, and SQL injection. The training and testing used in the XGBoost model was based on a CVE dataset obtained from the National Vulnerability Database where major features impacting exploitability were targeted. Proposed model obtained good model accuracy, precision, recall, and an F1-score, of 94%, 92%, 91%, and 92%, respectively, thereby proving to be a very efficient model in distinguishing between exploitable and non-exploitable vulnerabilities. This work contributes to the domain of vulnerability analysis by presenting the potential of XGBoost on CVE classification, along with several actionable insights into vulnerability management.

关键词： Training Accuracy Machine learning algorithms Operating systems Linux SQL injection Vectors Software Cryptography Testing

来源：评论

学校读者我要写书评

暂无评论

An Integrated Platform for Virtual Ecosystems and Wireless Edge Computing for Web 3.0 3

An Integrated Platform for Virtual Ecosystems and Wireless E...

引用

3rd IEEE International Conference on Intelligent Techniques in Control, Optimization and Signal Processing, INCOS 2024

作者： Garg, Khushi Jattu, Mahee Hariharasitaraman, S. Subramanian, R. Raja Division of Cyber Security and Digital Forensics VIT Bhopal University Madhya Pradesh India School of Computing Science and Engineering Vit Bhopal University Madhya Pradesh India Department of Computer Science and Engineering Kalasalingam Academy of Research and Education Tamil Nadu India

ISBN: (纸本)9798350361186

The emergence of Web 3.0 has paved the way for a new generation of the internet that enables users to view, publish, and own content independently. This innovation is being driven by a combination of blockchain, semantic communication, edge computing, and artificial intelligence, which together create value networks that facilitate participatory decision-making. Blockchain, in particular, offers security services by recording content in a decentralized and transparent manner. Additionally, it is compatible with decentralized wireless edge computing designs that can accurately express the intended meanings of contents without requiring a lot of resources. Building on this foundation, this research proposes a paradigm for unified blockchain-semantic ecosystems for Web 3.0 with wireless edge intelligence. Specifically, we present a semantic proof technique built on Oracle that enables Web 3.0's off-chain and on-chain interactions. To improve interaction efficiency and support Web 3.0, Oracle has developed an adaptable Deep Reinforcement Learning-based sharding technique. Our work also introduces a blockchain-based semantic exchange architecture that tokenizes semantic data into non-fungible tokens (NFT) for semantic exchange in this system. We employ Zero-Knowledge Proof to trade real semantic information without publicly posting it before accepting payment, in contrast to traditional NFT marketplaces. This makes it possible to trade fairly and privately. Overall, our proposed framework can precisely perform decentralized semantic sharing and information transfer, further achieving the advantages of semantic extraction and communication in Web 3.0. © 2024 IEEE.

关键词： Edge computing

来源：评论

学校读者我要写书评

暂无评论

CryptoPyt: Unraveling Python Cryptographic APIs Misuse with Precise Static Taint Analysis 40

CryptoPyt: Unraveling Python Cryptographic APIs Misuse with ...

引用

40th Annual computer security Applications Conference, ACSAC 2024

作者： Guo, Xiangxin Jia, Shijie Lin, Jingqiang Ma, Yuan Zheng, Fangyu Li, Guangzheng Xu, Bowen Cheng, Yueqiang Ji, Kailiang School of Cyber Security University of Science and Technology of China Hefei China Key Laboratory of Cyberspace Security Defense Institute of Information Engineering CAS Beijing China School of Cryptology University of Chinese Academy of Sciences Beijing China Faculty of Electrical Engineering and Computer Science Ningbo University Ningbo China NIO China

ISBN: (纸本)9798331520885

Cryptographic APIs are essential for ensuring the security of software systems. However, many research studies have revealed that the misuse of cryptographic APIs is commonly widespread. Detecting such misuse in Python poses challenges due to its intricate features, including dynamic features and pass-by-object-reference. Existing tools lack the precision and accuracy to tackle these challenges, leading to both high false positives and false negatives. In this work, we propose a specific Python Cryptographic Abstract Syntax Tree (PCAST) to represent the structure of source code, which rewrites AST nodes to handle complex Python features. Based on PCAST, we design and implement CryptoPyt, a static code analysis tool that leverages precise taint analysis and 17 cryptographic misuse rules to automatically identify potential cryptographic APIs misuse in Python projects. We conduct an in-depth analysis of all the APIs within the popular 21 Python cryptographic libraries and design five kinds of taint detectors to perform intra-procedural and inter-function analysis on the APIs and arguments. To demonstrate the effectiveness of CryptoPyt, we conduct experiments with six state-of-the-art tools (i.e., Cryptolation, LICMA, Bandit, Dlint, Semgrep and CodeQL) on both the labeled benchmark PyCryptoBench and the real-world Python cryptographic projects datasets PCAMD. Our evaluations show that CryptoPyt achieves an F1 score of 0.80 on PyCryptoBench and a recall rate of 99.08% on PCAMD. Furthermore, we disclose the discovered critical issues to the developers and seven high-level CVE IDs have been assigned to these findings. Our tool contributes to enhancing the security of Python cryptographic software. © 2024 IEEE.

关键词： Python

来源：评论

学校读者我要写书评

暂无评论

Differentially Private and Heterogeneity-Robust Federated Learning With Theoretical Guarantee

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第12期5卷 6369-6384页

作者： Wang, Xiuhua Wang, Shuai Li, Yiwei Fan, Fengrui Li, Shikang Lin, Xiaodong Huazhong University of Science and Technology Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security School of Cyber Science and Engineering Wuhan430074 China University of Electronic Science and Technology of China National Key Laboratory of Wireless Communications Chengdu611731 China Xiamen University of Technology Fujian Key Laboratory of Communication Network and Information Processing Xiamen361024 China University of Guelph School of Computer Science GuelphONN1G 2W1 Canada

Federated learning (FL) is a popular distributed paradigm where enormous clients collaboratively train a machine learning (ML) model under the orchestration of a central server without knowing the clients' private raw data. The development of effective FL algorithms faces multiple practical challenges including data heterogeneity and clients' privacy protection. Despite that numerous attempts have been made to deal with data heterogeneity or rigorous privacy protection, none have effectively tackled both issues simultaneously. In this article, we propose a differentially private and heterogeneity-robust FL algorithm, named DP-FedCVR to mitigate the data heterogeneity by following the client-variance-reduction strategy. Besides, it adopts a sophisticated differential privacy (DP) mechanism where the privacy-amplified strategy is applied, to achieve a rigorous privacy protection guarantee. We show that the proposed DP-FedCVR algorithm maintains its heterogeneity-robustness though DP noises are incorporated, while achieving a sublinear convergence rate for a nonconvex FL problem. Numerical experiments based on image classification tasks are presented to demonstrate that DP-FedCVR provides superior performance over the benchmark algorithms in the presence of data heterogeneity and various DP privacy budgets. © 2020 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Bridging Visual Affective Gap: Borrowing Textual Knowledge by Learning from Noisy Image-Text Pairs 24

Bridging Visual Affective Gap: Borrowing Textual Knowledge b...

引用

32nd ACM International Conference on Multimedia, MM 2024

作者： Wu, Daiqing Yang, Dongbao Zhou, Yu Ma, Can School of Cyber Security University of Chinese Academy of Sciences Beijing China Institute of Information Engineering Chinese Academy of Sciences Beijing China TMCC College of Computer Science Nankai University Tianjin China

ISBN: (纸本)9798400706868

Visual emotion recognition (VER) is a longstanding field that has garnered increasing attention with the advancement of deep neural networks. Although recent studies have achieved notable improvements by leveraging the knowledge embedded within pre-trained visual models, the lack of direct association between factual-level features and emotional categories, called the ''affective gap'', limits the applicability of pre-training knowledge for VER tasks. On the contrary, the explicit emotional expression and high information density in textual modality eliminate the ''affective gap''. Therefore, we propose borrowing the knowledge from the pre-trained textual model to enhance the emotional perception of pre-trained visual models. We focus on the factual and emotional connections between images and texts in noisy social media data, and propose Partitioned Adaptive Contrastive Learning (PACL) to leverage these connections. Specifically, we manage to separate different types of samples and devise distinct contrastive learning strategies for each type. By dynamically constructing negative and positive pairs, we fully exploit the potential of noisy samples. Through comprehensive experiments, we demonstrate that bridging the "affective gap'' significantly improves the performance of various pre-trained visual models in downstream emotion-related tasks. Our code is released on https://***/wdqqdw/PACL. © 2024 Owner/Author.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Circumventing Google Play vetting policies: a stealthy cyberattack that uses incremental updates to breach privacy

引用

Journal of Ambient Intelligence and Humanized Computing 2023年第5期14卷 4785-4794页

作者： Muhammad, Zia Amjad, Faisal Iqbal, Zafar Javed, Abdul Rehman Gadekallu, Thippa Reddy Department of Computer Science North Dakota State University Fargo United States Department of Cyber Security PAF Complex E-9 Air University Islamabad Pakistan Islamabad Pakistan School of Information Technology and Engineering Vellore Institute of Technology Vellore India Department of Electrical and Computer Engineering Lebanese American University Byblos Lebanon

Today digital technologies are evolving to accommodate small businesses and young entrepreneurs by reducing their time-to-market while encouraging rapid innovation in mobile, Extended Reality (XR), Internet of Things (IoT), cloud, and edge devices. The leading operating system Android typically takes one to a few days to perform application vetting and go to production by leveraging code analysis technologies in their Play Protect anti-malware program. However, developers with malicious intent are looking to circumvent this detection mechanism by exploiting Google’s relatively lenient trust policies that allow for package distribution and feature updates. This paper develops a proof-of-concept malware that exploits customers’ trust and Google’s policies to circumvent popular voice search applications. Our results show that attackers can initially circumvent Play Protect by uploading benign applications to build trust and then add malicious feature updates incrementally to distribute highly intrusive malware into user systems. This malware can scan and collect private user data from the device and exfiltrate it to the command-and-control server. The contributions are three-fold. (1) A proof-of-concept stealthy malware and publishing mechanism has developed that highlights the relative ease with which Google Play Protect policies may be subverted. (2) a comprehensive evaluation has been performed using major publicly available anti-malware solutions. (3) Recommendations and policies have been suggested to prevent this attack and ensure users’ privacy concerns (IMUTA is a novel attack in which malicious functionality is slowly added to a benign application through updates. This attack evades malware detection tools and exploits user trust. The attack can be launched against any application distribution platform like the Play Store). © 2023, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.

关键词： Mobile security

来源：评论

学校读者我要写书评

暂无评论

Intelligent Deep Learning Model for Securing KMS Through CSF Deviation Detection 13

Intelligent Deep Learning Model for Securing KMS Through CSF...

引用

13th International Conference on System Modeling and Advancement in Research Trends, SMART 2024

作者： Chauhan, Raj Ahmad, Shahnawaz Onyarin, Ogala Justin Kumar, Rajendra Noida Institute of Engineering and Technology Department of CSE-AIML Greater Noida India School of Computer Science Engineering and Technology Bennett University Greater Noida India University of Delta Cyber Security Department Delta State Nigeria Jamia Millia Islamia Department of Computer Science New Delhi India

ISBN: (纸本)9798350380583

Cloud computing has revolutionized data content stock and administration but concurrently it has increased the vulnerability factors especially Key Management System (KMS), which is crucial since it is used to protect cypher key. Typically, there are several limitations to anomaly detection methods used currently: poor precision and insufficient speed to detect threats. This research presents an intelligent model that integrated deep learning technique into existing Cloud security Framework (CSF) to conduct proactive identification on the deviation from the baseline behavior of interactions between the applications and the KMS. This is achieved through continually accruing KMS log data, defining behavioral profiles for the applications and using clustering to categorize them appropriately. The model also continuously monitors the system for any deviations, making anomalies easily discernible for quick action either through raising alerts to stakeholders or when using the KMS API to automatically respond to the flagged issues. Studies in the multitenant cloud scenario indicate that implementation of LSTMbased model is more effective than conventional approaches, enhancing the CSF's capability in threat detection thus making it possible to quickly mount defensive measures to safeguard crucial data and cryptographic keys. © 2024 IEEE.

关键词： Cloud security

来源：评论

学校读者我要写书评

暂无评论

Automated Crop Growth Monitoring and Optimizing the Yield with Reinforcement Learning 3

Automated Crop Growth Monitoring and Optimizing the Yield wi...

引用

3rd IEEE International Conference on Mobile Networks and Wireless Communications, ICMNWC 2023

作者： Gorantla, Snehitha Veluchamy, S. Amrita School of Computing Department of Computer Science and Engineering - Artificial Intelligence Amrita Vishwa Vidyapeetham Chennai601103 India Amrita School of Computing Department of Computer Science and Engineering - Cyber Security Amrita Vishwa Vidyapeetham Chennai601103 India

ISBN: (纸本)9798350317022

A reinforcement learning agent for optimal green-house management through Proximal Policy Optimisation is developed and evaluated in this work. With variable elements like lighting, irrigation, humidity, and temperature, a virtual greenhouse is produced. By interacting with the surroundings, the agent modifies the controls and gets input on things like crop health and growth rate. For policy and value estimation, the neural network design consists of distinct actor and critic networks as well as shared feature layers. Compared to fixed or random policies, the trained agent shows the ability to regulate conditions for increased crop yield and sustainability. As the number of trials increased, the RL agent's average return per trial increased steadily. This suggests that the RL agent has become more adept at controlling the greenhouse environment in subsequent experiments. This demonstrates how well the RL agent can manage crop health and resource use in the greenhouse climate control system. © 2023 IEEE.

关键词： Crop Yield Greenhouse Neural Network PPO algorithm Reinforcement 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：