检索结果-内蒙古大学图书馆

International Conference on Dependable systems and Networks Workshops (DSN-W)

作者： Ao Ding Gaolei Li Xiaoyu Yi Yuchen Liu School of Electronic Information and Electrical Engineering Shanghai Jiao Tong University China Shanghai Key Laboratory of Integrated Administration Technologies for Information Security China Department of Computer Science North Carolina State University USA

Software vulnerability detection is evolving from pattern-driven methods to data-driven methods to be more automatic and intelligent due to the emergence of deep learning. However, false positives and coarse-grained detection can not tackle changeable and unknown threats (e.g., Oday vulnerabilities), and also low model interpretability leads to incontinent detection results or maliciously being manipulated. To enhance model confidence and interpretability, in this paper, we propose an Interpretable vulnerability signature (IVSign) scheme to combine pattern-driven and data-driven methods. In IVSign, pattern-driven methods utilize static analysis to provide rich vulnerability trace information extracted from data dependencies and data-driven methods respectively train a transformer-based model to encode these vulnerability traces and a recurrent neural network to generate vulnerability signatures. In the testing stage of IVSign, if a program sample with vulnerabilities is fed into the pipeline, its hit rate in the signature database will be very high. Experiments have shown that when the matching threshold is set appropriately, the evaluation metrics on the testing samples exceed 90.00%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A-FAR: Research on Anti-obscuring Face Recognition Based on Improved FaceNet Model

A-FAR: Research on Anti-obscuring Face Recognition Based on ...

引用

International Conference on Advanced Cloud and Big Data (CBD)

作者： Biaokai Zhu Yangyang Zhang Lu Li Yuhong He Yiting Song Xiafei Li Ruiting Yin Jie Song Ping Li Sanman Liu Jumin Zhao Cyber Security Department Shanxi Police College Taiyuan University of Technology Taiyuan China Cyber Security Department Shanxi Police College Taiyuan China Intelligent Policing Key Laboratory of Sichuan Province Sichuan Police College Luzhou China School of Computer Science and Technology Anhui University Hefei China College of Information and Computer Taiyuan University of Technology Jinzhong China

Face recognition and cloud computing are the current hot research areas, in real life, face image acquisition is often affected by various factors, such as occlusion, lighting and facial posture, many application scenarios require in the occlusion environment to be able to achieve face recognition. The objective of this paper is to realize a system for collecting and preprocessing face images, processing, analyzing and recognizing them on the cloud platform. An occluse-resistant face recognition algorithm based on the improved FaceNet model is proposed. The system uses MTCNN for face detection and the improved FaceNet for face recognition. After a series of experiments, the system performs well in the evaluated data set, with an accuracy rate of 94.76%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Halftone Image Steganography Based on Maximizing Visual Similarity of Block Units 7th

Halftone Image Steganography Based on Maximizing Visual Sim...

引用

7th International Conference on Artificial Intelligence and security, ICAIS 2021

作者： Yin, Xiaolin Yu, Mujian Chen, Lili Wang, Ke Lu, Wei School of Computer Science and Engineering Guangdong Key Laboratory of Information Security Technology Ministry of Education Key Laboratory of Machine Intelligence and Advanced Computing Sun Yat-sen University Guangzhou510006 China Guangdong Science and Technology Innovation Monitoring and Research Center Guangzhou510033 China

ISBN: (纸本)9783030786113

Steganography focuses on imperceptibility of both the human eyes but also to potential analyzers. Based on the human visual system (HVS) of halftone image, most previous visual quality measurements for multi-tone images can not be adopted to design the steganographic schemes. In this paper, we propose a halftone image steganographic scheme based on the maximizing visual similarity of block units aiming at improving the statistical security but also the image visual quality. First, the halftone HVS model is discussed. The similarity of a pixel and its neighboring regions (which are called block units) is an important factor for human eyes. Then the distortion score is designed based on the visual similarity for each pixel in an image. Several visual similarity scores owned by a pixel are obtained in different block units. In general, the maximal visual similarity score means that this pixel is optimal to embed data. Finally, these visual similarity scores are combined with a weight matrix, and the distortion score that measures the embedding distortion is sent into syndrome-trellis code (STC) to minimize the distortion. Experimental results demonstrate the proposed scheme preserves the visual quality and improves the statistical security. © 2021, Springer Nature Switzerland AG.

关键词： Pixels

来源：评论

学校读者我要写书评

暂无评论

Bidirectionally Guided Large Language Models for Consumer-Centric Personalized Recommendation

引用

IEEE Transactions on Consumer Electronics 2025年

作者： Yu, Linfang Xiao, Peng Xu, Liqun Kumari, Saru Alenazi, Mohammed J.F. Industrial Research Institute Chengdu610213 China University of Electronic Science and Technology of China Network and Data Security Key Laboratory of Sichuan Province Chengdu610054 China Chengdu University of Technology School of Computer and Network Security Chengdu China Industrial Research Institute Chengdu610213 China Chaudhary Charan Singh University Department of Mathematics Uttar Pradesh Meerut250004 India Department of Computer Engineering Riyadh11451 Saudi Arabia

With the increasing abundance of consumer electronics, consumers are facing the challenge of information overload, making personalized recommendation systems (RSs) crucial for enhancing consumer experience and promoting sales. Recently, LLM-driven RSs have attracted widespread attention. Nevertheless, the application of LLMs in recommendation still encounters challenges: how to effectively integrate user/item content (CT) and collaborative filtering (CF) information;how to effectively process heterogeneous token sequences;and how to efficiently generate recommendations. Therefore, we propose BGLLM4Rec, a bidirectionally guided LLM for recommendation. Specifically, a bidirectionally guided pre-training strategy is proposed to extend user/item ID tokens into the vocabulary of the pre-trained LLM, and leverage both CF LLM and CT LLM to learn their embedding representations. Meanwhile, a heterogeneous-homogeneous (Het-Hom) prompting strategy is proposed to decompose the input sequence into heterogeneous prompt and homogeneous main text parts, performing language modeling solely on the latter to enhance modeling effectiveness and stability. Finally, a recommendation-specific masked prompting fine-tuning strategy is employed to further refine the pre-trained CF LLM, enabling the efficient generation of multiple recommended items without hallucinations. Extensive experiments conducted on three public datasets have demonstrated that BGLLM4Rec exhibits exceptional performance in recommendation tasks. © 1975-2011 IEEE.

关键词： Collaborative filtering

来源：评论

学校读者我要写书评

暂无评论

Generating Personalized Imputations for Patient Health Status Prediction in Electronic Health Records

Generating Personalized Imputations for Patient Health Statu...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Yang Liu Weiyu Zhang Jiasheng Si Zhao Li Xueping Peng Wenpeng Lu Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center Qilu University of Technology (Shandong Academy of Sciences) Jinan China Shandong Provincial Key Laboratory of Computing Power Internet and Service Computing Shandong Fundamental Research Center for Computer Science Jinan China Evay Info Jinan China Australian Artificial Intelligence Institute University of Technology Sydney Sydney Australia

ISBN: (数字)9798350386226

ISBN: (纸本)9798350386233

Electronic health records (EHRs) play a crucial role in the development of personalized treatment plans for patients. However, EHRs are often highly incomplete, posing significant challenges for predictive modeling. While existing deep learning models employ various imputation techniques to reconstruct missing values, they often fail to represent missing data in a personalized manner and do not learn from the missingness patterns in EHRs data. This limitation reduces their effectiveness in practical personalized healthcare applications. To address this issue, we propose SPIME, a self-supervised model that generates personalized imputations in EHRs data for patient health status prediction. We introduce a personalized missing mask (PMM) based on the frequency of feature measurements. Additionally, we incorporate a masked imputation task (MIT) loss that minimizes the loss of artificially introduced missing values, thereby enhancing the model’s capability to handle missing data. SPIME adopts self-supervised pretraining to learn representations from personalized missing patterns and reconstructs missing data in the latent space. To further enhance representation learning, two independent attention mechanisms are applied separately across the feature and temporal dimensions. Experimental results on two real-world EHRs datasets show that SPIME outperforms existing state-of-the-art methods in predicting in-hospital mortality and decompensation, demonstrating its effectiveness in reconstructing missing data and predicting patient health status. The code will be published at https://***/cling6666/SPIME.

关键词： Training Representation learning Biological system modeling Mortality Predictive models Biomedical measurement Imputation Data models Frequency measurement Electronic medical records

来源：评论

学校读者我要写书评

暂无评论

Contrastive Learning-Based Generic Audio-to-Lyrics Alignment

Contrastive Learning-Based Generic Audio-to-Lyrics Alignment

引用

IEEE International Conference on Big Data and Cloud Computing (BdCloud)

作者： Zhenkun Jin Shuai Huang Xin Nie Xinlei Zhou Yuanyuan Yi Gefei Zhou Wuhan Business UIniversity Wuhan China Wuhan Research Institute of Posts and Telecommunications Wuhan China Hubei Key Laboratory of Distributed System Security Wuhan China Hubei Engineering Research Center on BigData Security Wuhan China School of Cyber Science and Engineering Huazhong University of Science and Technology Wuhan China School of Journalism and Information Communication Huazhong University of Science and Technology Wuhan China School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China

In this research, we tackled the complex task of aligning audio and lyrics content automatically. This task entails the precise matching of lyrics with the corresponding audio segments in songs, necessitating the coordination of two distinct modalities: text and audio. To surmount this challenge, we employed a contrastive learning approach that generates cross-modal embeddings capable of establishing connections between the audio and text domains. Furthermore, we adopted a hierarchical and cascading design. Initially, the system predicts synchronization at the sentence level and subsequently refines the alignment at the word level. This design equips the system to handle extended sequences effectively, as cross-correlations leverage secondary memory that corresponds to sequence length. Through this innovative approach, we developed a trainable end-to-end system that can harness weakly annotated training data, concurrently learning a robust text model optimized for alignment tasks. Our experiments demonstrated substantial enhancements in average alignment accuracy when utilizing our proposed system, underscoring its robustness compared to previous state-of-the-art models.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Cloud Based AIVR Training system for Emergency Rescue

Cloud Based AIVR Training System for Emergency Rescue

引用

International Conference on Advanced Cloud and Big Data (CBD)

作者： Meiya Dong Biaokai Zhu Jumin Zhao Jie Song School of Computer Science Peking University Beijing China Suzhou Industrial Technology Institute of Zhejiang University Suzhou China Cyber Security Shanxi Police College City Country College of Information and Optical Engineering Taiyuan University of Technology Taiyuan China Cyber security Intelligent Policing Key Laboratory of Sichuan Province Luzhou China

ISBN: (数字)9798331511074

ISBN: (纸本)9798331511081

This paper mainly discusses the current situation and existing problems of training exercises in emergency medicial scenarios, as well as the importance and necessity of developing an AI training system for emergency medical scenarios. First, we introduce the current status of emergency medical training and emphasize the potential value of AI technology in this field. Then, the goal of developing an AI training system is defined as improving the efficiency and quality of medical education and training. We summarize the advantages of AI technology in emergency medical exercises and make a review and selection analysis of its related technologies. In the design and implementation stage, we describe in detail the requirement analysis, function design, system architecture, and core module development of the emergency medical AI training system. After system integration and testing, we verified the feasibility and effect of the system. In the process of evaluation and optimization, we evaluated the performance and user experience of the system and put forward optimization strategies and security considerations for the problems. In the application case analysis, we collect user feedback to verify its efficiency and carry out an in-depth analysis of the advantages and disadvantages of the system. Finally, in the conclusion and prospect part, we summarize the research results, point out the existing problems and improvement directions and look forward to the future development trends and challenges of AI training systems for emergency medical scenarios.

关键词： Training Emergency medicine systems architecture system integration User experience security Time factors Artificial intelligence Optimization Testing

来源：评论

学校读者我要写书评

暂无评论

EeCA: A Novel Approach for Energy Conservation in MEC via NDN-Based Content Caching

EeCA: A Novel Approach for Energy Conservation in MEC via ND...

引用

Asia-Pacific Network Operations and Management Symposium (APNOMS)

作者： Jiaxin Xu Huiling Shi Haoxiang Chu Wei Zhang Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center (National Supercomputer Center in Jinan) Qilu University of Technology (Shandong Academy of Sciences) Jinan China

In the context of escalating demands on wireless networks owing to the widespread application of 5G and the Internet of Things (IoT), enhancing network performance and user experience through the convergence of Multi-access Edge Computing (MEC) and information-Centric Networking (ICN) has become a critical research focus. This paper proposes a novel Energy-efficient Caching Algorithm (EeCA) based on Named Data Networking (NDN), a specific implementation of ICN. This NDN-based caching strategy extends local caching in MEC networks and enhances the overall network performance by considering the multitude of repeated computational tasks and the residual energy of low-energy devices. The EeCA aims to reduce redundant offloading of tasks across the entire network, contributing to energy savings, and reducing overall system costs. Simulations reveal superior performance of EeCA compared to four other caching methods, specifically in terms of cache hit ratio and energy consumption, making it a promising solution for network performance enhancement and energy conservation in MEC networks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Approximate state estimation with large-scale sensor networks under event-based sensor scheduling strategy

Approximate state estimation with large-scale sensor network...

引用

IEEE Conference on Decision and Control

作者： Xinhui Liu Meiqi Cheng Dawei Shi Ling Shi MIIT Key Laboratory of Servo Motion System Drive and Control School of Automation Beijing Institute of Technology Beijing China Department of Electronic and Computer Engineering Hong Kong University of Science and Technology Kowloon Hong Kong China

ISBN: (数字)9798350316339

ISBN: (纸本)9798350316346

In this paper, we investigate the issues of real-time sensor scheduling and state estimator design within large-scale sensor network systems. Specifically, data redundancy sometimes occurs in large-scale sensor arrays due to the excessive proximity of sensing units in the spatial domain, leading to high similarity in their measurement data. Currently, it is difficult to account for such redundancy in sensor scheduling algorithms found in existing literature, where the optimal subset of sensors is generally selected by optimizing objective functions formulated from certain performance criteria. To tackle this problem, we introduce an event-based sensor scheduling strategy, the triggering condition of which is designed founded on the similarity of sensor data, so as to identify the most informative subset of sensors for state estimation. To evaluate the impact of the sensor scheduling protocol on system observability, we propose a new notion of $\mathscr{E}(\varepsilon)$-observability, based on which an observability criterion is derived. In addition, we have designed a set-valued state estimation algorithm, which takes into account the intricate measurement information structure inherent within the sensor selection mechanism. The performance enhancement of the proposed estimator is also investigated. Finally, numerical experiments are conducted to validate the effectiveness of the proposed estimation algorithm and to verify the performance improvement.

关键词： Protocols Event detection Scheduling algorithms Redundancy Approximation algorithms Real-time systems Sensors State estimation Observability Sensor arrays

来源：评论

学校读者我要写书评

暂无评论

Identity Based Linkable Ring Signature with Logarithmic Size 17th

Identity Based Linkable Ring Signature with Logarithmic Size

引用

17th International Conference on information security and Cryptology, Inscrypt 2021

作者： Nassurdine, Mohamed Zhang, Huang Zhang, Fangguo School of Computer Science and Engineering Sun Yat-sen University Guangzhou510006 China Guangdong Province Key Laboratory of Information Security Technology Guangzhou510006 China School of Computer and Communication Engineering Changsha University of Science and Technology Changsha410114 China

ISBN: (纸本)9783030883225

Anonymity is an inevitable and sensitive matter of concern, especially in the age where people are willing to use digital devices and Internet to deal with almost all things in their work and daily life. To support anonymity, modern cryptography is one of the most suitable choices in the algorithm level. Particularly, in the scenarios, such as e-voting, crypto-currency, and smart grid etc., a cryptographic primitive, called linkable ring signature, has shown its ability to handle anonymity problems. However, signature schemes will introduce additional costs to those e-commerce systems, so that they should be as efficient as possible. On the other side, a signature scheme that requires the public key infrastructure (PKI) brings much unnecessary inconvenience to its users, since typically, users of the aforementioned systems are not familiar with cryptographic skills and the system establishers are trusted by them in some sense. As a result, an identity-based (ID-based) signature scheme with small size fulfills the visible requirements of e-commerce systems. In this paper, we proposed an ID-based linkable ring signature scheme with logarithmic size from pairing and elliptic curve discrete logarithm problem (ECDLP), and gave all the security proofs in detail. Besides that, the scheme needs no trusted setup, except that the key generation center knows the secret key of each user and it is a property of ID-based cryptography in nature. © 2021, Springer Nature Switzerland AG.

关键词： Digital devices

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：