检索结果-内蒙古大学图书馆

2025 ieee international Students' conference on Electrical, Electronics and computer Science, SCEECS 2025

作者： Srivastava, Saurabh Nakhate, Sangeeta Maulana Azad National Institute of Technology Dept. of Ece Bhopal India

ISBN: (纸本)9798331529833

System security is becoming an increasingly essential feature of modern computing systems, particularly programmable hardware such as Field Programmable Gate arrays (FPGA), employed in various applications such as autonomous systems and digital networks. However, safeguarding FPGA-based devices against IP theft and reverse engineering is difficult and a big concern for the semiconductor and defense industries. Previously, various obfuscation approaches such as gate level obfuscation, camouflaging, and strong puff were employed to make a distinct key for every chip. These designs, however, are more area-intensive, consume more power, and are easier to crack the key. In this research, we propose a novel obfuscation idea that integrates a reconfigurable hardware obfuscation block into the original circuit with an additional N-stage Delay circuit to enhance overall system security and resistance to threats. The present idea relates to the increase in the security of the system against the piracy, overbuilding and from the cloning and camouflaging of the chip. The proposed technique does not have any bulky circuit, and it is easily realizable and just requires some additional modules that can easily reconfigured with a very smaller area and power overhead thereby offering performance benefits and increasing the time of sat attack iteration, making the locking system more complex and difficult to decode. © 2025 ieee.

关键词： Reverse engineering

来源：评论

学校读者我要写书评

暂无评论

Comparative Analysis of Traditional and Ensemble Learning Models for Network Security System

Comparative Analysis of Traditional and Ensemble Learning Mo...

引用

2025 ieee international conference on Computational, Communication and Information Technology, ICCCIT 2025

作者： Verma, Anand Rathod, Maya Oriental University Department of Computer Science and Engineering Indore India

ISBN: (纸本)9798331512965

In this paper use CIC-IDS2017 dataset to illustrate a comparative analysis of traditional and proposed models for intrusion detection in network security systems. The comparison includes DT, RF, ET and XGBoost classifiers. Normalize class, SMOTE Data preprocessing to compensate for missing values and of imbalance classes From the results of evaluation metrics, we conclude that proposed methods show a significant improvement on its model performances in detecting network intrusions as compared to traditional models XGboost specifically this work. Out of all the algorithms, the accuracy for XGBoost stands out, as it is mostly known to cater to complex attack patterns. The results indicate that ensemble base techniques offer the best solutions which can prove to be both effective and efficient as an intrusion detection systems for current network technologies. © 2025 ieee.

关键词： Network intrusion

来源：评论

学校读者我要写书评

暂无评论

Graph Representation Learning for Recommender systems

Graph Representation Learning for Recommender Systems

引用

2025 ieee international Students' conference on Electrical, Electronics and computer Science, SCEECS 2025

作者： Manikrao, Kulkarni Manjusha Hiremath, Savitha Manjula, S.H. Dayananda Sagar University School of Engineering Department of Computer Science and Engineering Karnataka Bengaluru India University of Visvesvaraya College of Engineering Department of Computer Science and Engineering Karnataka Bengaluru India

ISBN: (纸本)9798331529833

Graph data represents information efficiently and can be used to learn subsequent tasks easily. In the domain of biological science, recommender systems, social network analysis graph representation learning has become ubiquitous with tremendous success. A graph structured data can make a system to learn, produce new information and can be analyzed for set of tasks. Deep learning-based recommender systems are using deep graph embeddings for improved accuracy by exploring user item relations. This relation exploration can be utilized for other recommendation task such as link prediction. In this study we are providing overview of graph representation learning with complete analysis of data with graph representation. This study provides importance of graph representation in recommender system to learn node representations. The results and discussion section denotes how simple graph neural network technique successfully applied on graph data for node representation learning and challenges. © 2025 ieee.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

On the Misspecified Cramér-Rao Bound and Elliptically Symmetric Distributions 59

On the Misspecified Cramér-Rao Bound and Elliptically Symme...

引用

59th Annual conference on Information Sciences and systems, CISS 2025

作者： Richmond, Christ D. Duke University Pratt School of Engineering Department of Electrical and Computer Engineering DurhamNC United States

ISBN: (纸本)9798331513269

Parameter estimation is among the many goals of engineering systems and scientific applications, and parameter bounds have provided significant insights into achievable system performance. These bounds are based on an assumed distribution for the data that may differ from the true data distribution, i.e. often the impacts of model misspecification are inevitable. Thus, researchers have derived Slepian-Bangs-type formulas for elliptically symmetric distributions under model misspecification to explore these effects. The present work continues this exploration by establishing a simple version of the Slepian formula applicable to circular complex compound Gaussian distributed data (a special class of non-Gaussian distributions, also sometimes called Gaussian mixtures), when the data distribution assumed for algorithm development is complex Gaussian. © 2025 ieee.

关键词： Cramer-Rao bounds

来源：评论

学校读者我要写书评

暂无评论

Overview of Accent Conversion Techniques Based on Deep Learning 6

Overview of Accent Conversion Techniques Based on Deep Learn...

引用

6th ieee international conference on Image Processing, Applications and systems, IPAS 2025

作者： Shrestha, Aayam Ghorshi, Seyed Panahi, Issa The University of Texas at Tyler Department of Electrical and Computer Engineering TylerTX United States The University of Texas at Dallas Department of Electrical and Computer Engineering RichardsonTX United States

ISBN: (纸本)9798331506520

In modern communication, speech technology plays a vital role, with accent being one of the key features that reflects the nuances of spoken technology. However, differences in accent can make communication difficult, especially in global contexts like language learning and human computer interaction. Accent conversion approaches often solve these issues by transforming the accent of speaker to that of the other speakers, while maintaining linguistic information and speaker identity. This technology holds great applications to speech recognition, language learning, and cross-lingual communication. This paper explores the use of advanced deep learning techniques in accent conversion systems which have improved both the naturalness and intelligibility of converted speech. Additionally, we discuss the methods generally used for performance evaluation of these systems. © 2025 ieee.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Advanced Facial Emotion Classification with 135 Classes for Enhanced Cybersecurity Applications 4

Advanced Facial Emotion Classification with 135 Classes for ...

引用

4th ieee international conference on Artificial Intelligence in Cybersecurity, ICAIC 2025

作者： Powers, Gregory Jin, Aobo Tonmoy, Abdul Basit Cao, Haowei Gohel, Hardik Deng, Qixin Wabash College Mathematics and Computer Science Crawfordsville United States University of Houston Victoria Computer Science Victoria United States Johns Hopkins University Electrical and Computer Engineering Baltimore United States

ISBN: (纸本)9798331518882

A key component of behavior analysis and human-computer interaction (HCI) is facial expression detection, which helps systems understand and react to human emotions more effectively and nuancedly. While previous research predominantly focuses on classifying seven or eight primary emotions, there has been limited exploration into more detailed and varied emotional expressions. In this paper, we address this gap by utilizing a recently published dataset that includes 135 distinct emotional classes, significantly increasing the complexity and granularity of emotion recognition tasks. To tackle this challenge, we propose a novel framework that combines U-Net and ResNet architectures, leveraging the strengths of each to capture intricate emotional cues and spatial features. Our approach achieves higher accuracy than current state-of-the-art (SOTA) methods, demonstrating its effectiveness for high-dimensional emotion classification in complex, real-world applications. © 2025 ieee.

关键词： computer vision deep learning facial emotion recognition

来源：评论

学校读者我要写书评

暂无评论

Diagnosability Verification and Enforcement for Unbounded Petri Nets by Online Supervisors

引用

ieee TRANSACTIONS ON AUTOMATION SCIENCE AND engineering 2025年 22卷 9061-9074页

作者： Hu, Shaopeng Hu, Yihui Liu, Ding Fanti, Maria Pia Li, Zhiwu Xidian Univ Sch Electromech Engn Xian 710071 Peoples R China Xian Univ Posts & Telecommun Sch Automat Xian 710121 Peoples R China Polytech Univ Bari Dept Elect & Informat Engn I-70125 Bari Italy Macau Univ Sci & Technol Inst Syst Engn Taipa Macao Peoples R China

This paper addresses the problems of diagnosability verification and enforcement of discrete event systems modeled with unbounded Petri nets. Diagnosability in such systems is critical for ensuring reliability and maintaining operational integrity, yet current methods often struggle with the complexity introduced by unboundedness and potential deadlocks. Given an unbounded labeled Petri net that may reach deadlocks, a quiescent basis coverability graph is established to verify the diagnosability of the considered system. This procedure employs a deterministic finite state automaton, called an extended verifier, derived from the proposed quiescent basis coverability graph. It is shown that an unbounded Petri net is diagnosable if and only if the verifier does not contain a class of cycles, called repetitive $F$ -cycles. This result also provides necessary and sufficient conditions for diagnosability enforcement by developing an online supervisor. Further, the designed supervisor is maximally permissive and also circumvents a plant entering deadlocks by firing non-fault sequences. Examples are presented to demonstrate the proposed method. Note to Practitioners-Fault diagnosis and diagnosability enforcement are critical for the development and operation of highly automated systems covering computer-integrated production processes, intelligent traffic, computer and communication networks, smart gird, etc. This work touches upon this problem from the perspective of discrete event systems that are modeled with unbounded labeled Petri nets. The feasibility and applicability of the reported method stem from the usage of a structurally compact representation of a considered plant such that the computational cost of a real-world system is acceptable. The graphical representation of Petri nets as well as the proposed quiescent basis coverability graph make the method easy to use and manipulate. Moreover the sufficient and necessary conditions of diagnosability enforcement can

关键词： Petri nets System recovery Fault diagnosis Automation Automata Discrete-event systems Polynomials Mechanical engineering Firing Computational modeling Unbounded Petri net fault diagnosis diagnosability enforcement discrete event system

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Study of Case-Based Reasoning in Information Retrieval: An Artificial Intelligence Approach

A Comprehensive Study of Case-Based Reasoning in Information...

引用

2025 international conference on Pervasive Computational Technologies, ICPCT 2025

作者： Solanki, Surabhi Verma, Seema Biswas, Prolay Roy, Pulakesh Singh, Aparna Sharma, Surbhi Computer Science & Engineering Banasthali Vidyapith Rajasthan Tonk India Scset Bennett University Greater Noida India Education National Institute of Technical Teachers' Training and Research Department of Electrical & Electronics Engineering Bhopal India Computer Science & Engineering Dayananda Sagar University Ramanagara India Computer Science& Engg Brainware University Kolkata India

ISBN: (纸本)9798331508685

Information Retrieval (IR) systems which provide accurate and contextually relevant information to healthcare and law etc. where accuracy is very crucial. Traditional information retrieval methods fundamentally based on keyword matching and statistical models usually difficult to achieve the semantic meaning and contextual nuances of queries. Case-Based Reasoning (CBR), as an artificial intelligence technique, provide advance solutions to enhance IR systems based on a set of past cases through adaptive context-supporting retrieval. By taking advantage of past cases data, CBR allows IR systems to interpret complex context-sensitive queries and offer personalized results. The paper provides an overview of traditional IR models, examines the CBR integration into these systems, and evaluates the impact of CBR on retrieval effectiveness and personalization. And finally, we elaborate on the challenges involved with scaling and adapting cases towards real-time retrieval and provide some future directions for research, including integrating machine learning techniques and hybrid models to enhance the relevance, interpretability, and scalability of CBR-based IR systems. © 2025 ieee.

关键词： Case based reasoning

来源：评论

学校读者我要写书评

暂无评论

Research on Control Strategies for Partially Observable Discrete Event systems Based on Normal State Tree Structure 24

Research on Control Strategies for Partially Observable Disc...

引用

5th international conference on computer Science and Management Technology, ICCSMT 2024

作者： Gu, Chan Tang, Bo Chi, Encheng School of Electrical and Control Engineering Shaanxi University of Science and Technology Shaanxi Xi'an China

ISBN: (纸本)9798400709999

In the process of modeling and solving large-scale complex systems, issues with complex state structures and low computational efficiency are often *** partially observable discrete event systems, this paper proposes a method that utilizes state tree structures for modeling and analyzes global behavior through local *** demonstrates the existence of non-blocking, weakly controllable coreachable and observable (WCCO) subpredicates and their solving *** order to effectively represent these complex system states, the paper employs the use of Binary Decision Diagrams (BDD), which are capable of compactly and efficiently representing state tree *** feasibility of the proposed algorithm is verified through practical examples, showing that the algorithm exhibits higher fault tolerance and flexibility in handling complex systems, enabling the development of more forgiving control *** not only enhances the stability and reliability of the system but also provides new solutions for supervisory control in practical engineering applications. © 2024 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Binary decision diagrams

来源：评论

学校读者我要写书评

暂无评论

Attention-Driven Causal Discovery: From Transformer Matrices to Granger Causal Graphs for Non-Stationary Time-series Data

Attention-Driven Causal Discovery: From Transformer Matrices...

引用

2025 ieee international conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Zhu, Jiageng Li, Kehao Mai, Zheda Xie, Hanchen AbdAlmageed, Wael Abraham, Zubin Information Sciences Institute University of Southern California United States Sunnyvale United States Department of Computer Science and Engineering Ohio State University United States Holcombe Department of Electrical and Computer Engineering Clemson University United States

ISBN: (纸本)9798350368741

Causal discovery in non-stationary time series data is crucial for understanding complex systems but remains challenging due to evolving relationships over time. This paper presents a novel two-stage approach for causal discovery in non-stationary multivariate time series data. The first stage employs a Temporal Attention Forecasting Network (TAFNet), a modified Transformer architecture, to capture complex temporal dependencies and generate informative attention matrices. The second stage utilizes these matrices in an iterative process for Granger causality discovery, refining the predicted causal graph while improving forecasting accuracy. The proposed method addresses the limitations of existing approaches and provides a more complete understanding of causal relationships in non-stationary systems. Extensive experiments demonstrate the method's superior performance compared to state-of-the-art approaches, particularly in handling non-linear relationships and scaling to high-dimensional data. © 2025 ieee.

关键词： Causal Discovery Non Stationary Time Series

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：