检索结果-内蒙古大学图书馆

Joint 30th International Conference on Computational Linguistics and 14th International Conference on Language Resources and Evaluation, LREC-COLING 2024

作者： Wang, Di He, Yuzheng Liang, Xiao Tian, Yumin Li, Shaofeng Zhao, Lin School of Computer Science and Technology Xidian University Xi'an710071 China Guangzhou Institute of Technology Xidian University Guangzhou510555 China School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing210094 China

ISBN: (纸本)9782493814104

End-to-end multimodal aspect-based sentiment analysis (MABSA) combines multimodal aspect terms extraction (MATE) with multimodal aspect sentiment classification (MASC), aiming to simultaneously extract aspect words and classify the sentiment polarity of each aspect. However, existing MABSA methods have overlooked two issues: (i) They only focus on fusing image regional information and textual words for two subtasks of MABSA. Whereas, MATE subtask relies more on global image information to assist in obtaining the quantity and attributes of aspects. Ignoring the integration with global information may affect the performance of MABSA methods. (ii) They fail to take advantage of target information. Nevertheless, the fine-grained details of targets are important for classifying sentiments of aspects. To solve these problems, we propose a Target-oriented Multi-grained Fusion Network (TMFN). It fuses text information with global coarse-grained image information for MATE subtask and with fine-grained image information for MASC subtask. In addition, a target-oriented feature alignment (TOFA) module is designed to enhance target-related information in image features with target details. In such a way, image features will contain more target emotional-related information which is beneficial to sentiment classification. Extensive experiments show that our method outperforms state-of-the-art methods on two benchmark datasets. © 2024 ELRA Language Resource Association: CC BY-NC 4.0.

关键词： Sentiment analysis

来源：评论

学校读者我要写书评

暂无评论

Streamlined Data Pipeline for Real-Time Threat Detection and Model Inference 17

Streamlined Data Pipeline for Real-Time Threat Detection and...

引用

17th International Conference on COMmunication Systems and NETworkS, COMSNETS 2025

作者： Singh, Rajkanwar Aravindan, V. Mishra, Sanket Singh, Sunil Kumar VIT-AP University School of Computer Science & Engineering Andhra Pradesh Amaravati India Manipal Institute of Technology Bengaluru Manipal Academy of Higher Education Department of Computer Science and Engineering Karnataka Manipal India

ISBN: (纸本)9798331531195

Real-time threat detection in streaming data is crucial yet challenging due to varying data volumes and speeds. This paper presents an architecture designed to manage large-scale, high-speed data streams using deep learning and machine learning models. The system utilizes Apache Kafka for high-throughput data transfer and a publish-subscribe model to facilitate continuous threat detection. Various machine learning techniques, including XGBoost, Random Forest, and LightGBM, are evaluated to identify the best model for classification. The ExtraTrees model achieves exceptional performance with accuracy, precision, recall, and F1 score all reaching 99% using the SensorNetGuard dataset within this architecture. The PyFlink framework, with its parallel processing capabilities, supports real-time training and adaptation of these models. The system calculates prediction metrics every 2, 000 data points, ensuring efficient and accurate real-time threat detection. © 2025 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

An optimal group-wise non-rigid image registration technique based clinical decision support for dementia prediction

引用

Multimedia Tools and Applications 2025年 1-26页

作者： Mary, F. Sheeja Asanambigai, V. Fred, A. Lenin Assistant Professor Senior Grade School of Computing Dept. of CSE Vel Tech Rangarajan Dr. Sagunthala R & D Institute of Science and Technology Chennai India Computer Science and Engineering Annamalai University Chidambaram India Dept. of CSE Mar Ephraem College of Engineering and Technology Marthandam India

The fundamental challenge in clinical practice is predicting whether patients with cognitive symptoms or impairment will improve or stay stable. Population ageing is a worldwide phenomena with various impacts. Among dementia, Alzheimer disease (AD) is common in the elderly which is the most common form of dementia and may contribute to 60–70% of cases. An early and precise AD diagnosis enhances patient quality of life. In this study, we propose an optimal group-wise non-rigid image registration technique for dementia (AD) prediction. We introduce an improved cuckoo search optimization (ICSO) algorithm to search optimal group-wise dense correspondence in large image sets in non-rigid image registration which reduce dimensionality of the search space. Then, we illustrate a teacher-student learning based optimization algorithm which cooperates with PLS regression to extract deformation which separates the relationship information suitable for each other from the raw relationship information. After that, we compute the distribution of morphologies of age by fitting smooth percentile curves to these scores using hybrid capsule neural network with least Median Square (hybrid CNN-LMS) curve fitting model. Finally, the simulation results of proposed hybrid CNN-LMS technique are compared to the existing methods to analyze the measures. We observed that the accuracy of our proposed hybrid CNN-LMS technique is 96.23% and 95.82% for RSS and ANDI datasets respectively. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2025.

关键词： Neurodegenerative diseases

来源：评论

学校读者我要写书评

暂无评论

Removing Gaussian Image Noise Using Advanced Non-Local Means Filtering Techniques

Removing Gaussian Image Noise Using Advanced Non-Local Means...

引用

2023 International Conference on New Frontiers in Communication, Automation, Management and Security, ICCAMS 2023

作者： Suresh, A. Waris, Saiyed Faiayaz Mohankumar, B. Arumuga Arun, R. Saravanan, G. Vellore Institute of Technology School of Computer Science and Engineering Department of Computational Intelligence Vellore India Vignan's Foundation for Science Technology & Research Vadlamudi Department of Computer Science and Engineering Andhrapradesh Guntur India Vellore Institute of Technology School of Computer Science and Engineering Vellore India Veltech Rangarajan Dr.Sagunthala R&d Institute of Science and Technology Department of Computer Science and Engineering India

ISBN: (纸本)9798350317060

Our system is aimed at improving the application of image de-noising processes, even though the presence of noise has a negative impact on image quality. End-user applications suffer as a result of the focusing Gaussian noise. Introduce the Non-local means (NLM) filter to take advantage image similarities and analyse data preservation in de-noising. Even though the NLM approach eliminates some noises, like salt and pepper, an estimate weighted between the input images and edge preservation is still calculated. It could experience loss of visual structure, noise reduction, and blurring. This research shows the enhancement of the threshold level by combining stationary wavelet transform with NLM. This type of Advanced NLM method is effective at removing high noisy regions like Gaussian noise levels. The proposed ANLM method improves edge preservation and image structure similarities. This article uses some noisy images as input, then employs quantitative parameters to perform the measurement. Finally, this method demonstrates the effective removal noise and edge preservation of weighted refined structures. PSNR, COC, and MSE are the measuring parameters. © 2023 IEEE.

关键词： Mean square error

来源：评论

学校读者我要写书评

暂无评论

Resolving Attached USBs: Analysis of Windows 11 Artifacts 2

Resolving Attached USBs: Analysis of Windows 11 Artifacts

引用

2nd International Conference for Advancement in technology, ICONAT 2023

作者： Budhrani, Akash Singh, Upasna Singh, Bhupendra Defence Institute of Advanced Technology School of Computer Engineering and Mathematical Science Pune India Indian Institute of Information Technology Computer Science and Engineering Department Pune India

ISBN: (纸本)9781665475174

The Windows Operating System is known for its convenience which tends to breed more and more user information in form of Artifacts. Artifacts are important repository of potential evidence while conducting any computer investigation. This paper thus provides an analysis of artifacts that will bring out the inside of USB devices attached or mounted on a particular system. Moreover, this paper will parse the data out into usable information and helps to interpret the findings of forensic examiner. In this work, Image file of the suspected system has been exported from the system for further analysis to bring out relevant information. © 2023 IEEE.

关键词： Windows operating system

来源：评论

学校读者我要写书评

暂无评论

Universal Face Manipulation Proactive Defense by Frequency-Driven Imperceptible Adversarial Attack 18th

Universal Face Manipulation Proactive Defense by Frequency-...

引用

18th Chinese Conference on Biometric Recognition, CCBR 2024

作者： Wen, Jinchang Wang, Jian Sun, Yunlian Tistarelli, Massimo School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing China School of Computer Science Nanjing University of Posts and Telecommunications Nanjing China Department of Biomedical Sciences and Information Technology University of Sassari Sassari Italy

ISBN: (纸本)9789819610709

Currently, highly realistic face-swapped videos/images pose significant potential security risks to society, which forces researchers to develop an effective defense mechanism against deepfakes, especially in a proactive manner. This involves adding imperceptible adversarial perturbations to facial images, which render deepfake models ineffective. Although existing proactive defense methods based on adversarial perturbations show promising performance in white-box scenarios, most of them suffer from significant performance degradation in black-box scenarios. Moreover, we observe that adversarial perturbations appearing in low-frequency areas (e.g., cheeks) are more easily noticed by the human eye. To this end, we propose a novel Frequency-Aware Multi-Teacher Distillation (FMTD) framework, which aims to generate a universal and imperceptible perturbation with remarkable generalization property against various facial manipulation methods. To extract universal property, we propose a multi-teacher distillation block for learning valuable shared knowledge from different perturbation generators. Moreover, we design a Frequency-aware Mean Squared Error (FMSE) loss based on wavelet transform to prefer to generate imperceptible perturbations in high-frequency regions. Experimental results demonstrate the superior performance of our proposed proactive defense method to state-of-the-art methods. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Mean square error

来源：评论

学校读者我要写书评

暂无评论

Investigating the Impact of Public-Key Infrastructure in Information Security Models for Networked Applications 5th

Investigating the Impact of Public-Key Infrastructure in Inf...

引用

5th International Conference on Data science, Machine Learning and Applications, ICDSMLA 2023

作者： Shree, Ritu Saxena, Kumud Mehta, Deepak Pandeya, Megha Department of Computer Science and Application Vivekananda Global University Jaipur India Department of Computer Science Engineering Noida Institute of Engineering and Technology Uttar Pradesh Greater Noida India Karnataka Bangalore India Maharishi School of Engineering and Technology Maharishi University of Information Technology Uttar Pradesh Lucknow India

ISBN: (纸本)9789819780426

Public-key infrastructure (PKI) is poised to play a crucial function in the use of secure data technology systems. With the increasing complexity of these dais’s networks and packages, PKI gives a reliable way to electronically affirm and securely exchange critical records. This paper investigates the effect of PKI on modern-day facts security models for networked applications. An assessment of the literature is conducted to discover the limitations of various PKI‐enabled architectures and packages. The research seems at the usage of cryptography and digital certificates inside the implementation of PKI, and evaluates the different varieties of authentication, authorization, and encryption strategies in shielding networked structures. Furthermore, the paper discusses the advantages and demanding situations associated with the deployment of PKI for distinctive use instances. The findings exhibit that during order for PKI to efficiently defend networked structures, a sound authentication system need to be blended with sturdy and consumer-friendly encryption strategies. Moreover, the paper shows strategies for overcoming the demanding situations related to enforcing PKI in a relaxed and efficient way. Public-key infrastructure (PKI) is an important device in data security fashions for networked packages. PKI is a security architecture that entails the use of public and private interchangeable cryptographic keys to ensure authenticated and comfortable verbal exchange among entities. Through developing a hierarchical accept as true with device, virtual certificate are used to cozy communications and the sender and authenticity and integrity of records. This architecture gives a at ease foundation for at ease dispensed computing across multiple sites located in numerous geographical places, as it enhances confidentiality, integrity, and authentication of exchanged information. In addition, PKI can facilitate cozy statistics sharing, alleviate potential privateness issues,

关键词： Authentication

来源：评论

学校读者我要写书评

暂无评论

CMOS-Based Ripple Correlation Control MPPT DC-DC Boost Converter for Perovskite Solar CellEnergy Harvester 4

CMOS-Based Ripple Correlation Control MPPT DC-DC Boost Conve...

引用

4th IEEE International Conference in Power engineering Applications, ICPEA 2024

作者： Adebunmi, Kayode Oluwaseyi Marzuki, Arjuna Ng, Annie Ukaegbu, Ikechi Augustine Nazarbayev University School of Engineering and Digital Sciences Department of Electrical and Computer Engineering Astana Kazakhstan Wawasan Open University School of Science and Technology Penang Malaysia

ISBN: (纸本)9798350344578

The power management integrated circuits in photovoltaic (PV) energy harvesting applications require efficient maximum power point tracking (MPPT) algorithm to maximize power output from the PV in the face of continuously fluctuating atmospheric circumstances. A CMOS-based Ripple Correlation Control (RCC) algorithm for optimizing power extraction from the Perovskite Solar cell (PSC) was designed in this work. The RCC uses the intrinsic ripple characteristics of the DC-DC boost converter to perturb the voltage and current of the PV panel to achieve MPP. The design uses a multiplier, a 2-stage op-amp, two differentiator circuits, an integrator, an XOR gate, and a deadtime circuit. The PWM stage of the MPPT consists of a comparator, and a Schmitt trigger circuit was incorporated into the PWM section of the design to minimize the system's chip size and power consumption. The effectiveness of the RCC output signal was able to drive the boost converter to produces an output voltage of 5.67V from an input voltage of 1.5V with a power efficiency of 98%. The system was designed using UMC180nm CMOS technology. © 2024 IEEE.

关键词： Timing circuits

来源：评论

学校读者我要写书评

暂无评论

6G Non-Terrestrial Networks for Intelligent IoT Services

引用

IEEE Network 2024年第4期38卷 6-8页

作者： Jia, Min Chen, Hsiao-Hwa Chang, Zheng Zhang, Ning Wu, Zhibin Harbin Institute of Technology School of Electronics and Information Engineering National Cheng Kung University Department of Engineering Science University of Jyvaskyla Faculty of Information Technology University of Windsor Department of Electrical and Computer Engineering CETC Potevio Science & Technology Company Ltd.

Recently, the Internet of Things (IoT) has been considered for its applications in diverse settings, such as intelligent urban environments, advanced manufacturing facilities, and environmental surveillance. However, significant obstacles exist in the implementation of these IoT solutions due to various issues, such as heterogeneity, diverse application requirements, restricted resources, and massive connectivity. To tackle these issues, merging satellite-based networks with conventional ground-based networks has been proposed as an effective and affordable option for enhancing IoT device connectivity.

关键词： Special issues and sections 6G mobile communication Internet of Things Security Sensors Semantics Resource management computer architecture Collaboration Intelligent systems

来源：评论

学校读者我要写书评

暂无评论

SEMDR: A Semantic-Aware Dual Encoder Model for Legal Judgment Prediction with Legal Clue Tracing

SEMDR: A Semantic-Aware Dual Encoder Model for Legal Judgmen...

引用

2024 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2024

作者： Liu, Pengjie Zhang, Wang Ding, Yulong Zhang, Xuefeng Yang, Shuang-Hua College of Science Northeastern University China School of Computer Science and Engineering Southern University of Science and Technology China University of Reading Department of Computer Science United Kingdom

ISBN: (纸本)9781665410205

Legal Judgment Prediction (LJP) aims to form legal judgments based on the criminal fact description. However, researchers struggle to classify confusing criminal cases, such as robbery and theft, which requires LJP models to distinguish the nuances between similar crimes. Existing methods usually design handcrafted features to pick up necessary semantic legal clues to make more accurate legal judgment predictions. In this paper, we propose a Semantic-Aware Dual Encoder Model (SEMDR), which designs a novel legal clue tracing mechanism to conduct fine-grained semantic reasoning between criminal facts and instruments. Our legal clue tracing mechanism is built from three reasoning levels: 1) Lexicon-Tracing, which aims to extract criminal facts from criminal descriptions;2) Sentence Representation Learning, which contrastively trains language models to better represent confusing criminal facts;3) Multi-Fact Reasoning, which builds a reasons graph to propagate semantic clues among fact nodes to capture the subtle difference among criminal facts. Our legal clue tracing mechanism helps SEMDR achieve state-of-the-art on the CAIL2018 dataset and shows its advance in few-shot scenarios. Our experiments show that SEMDR has a strong ability to learn more uniform and distinguished representations for criminal facts, which helps to make more accurate predictions on confusing criminal cases and reduces the model uncertainty during making judgments. All codes will be released via GitHub. © 2024 IEEE.

关键词： Semantics

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：