检索结果-内蒙古大学图书馆

4th International Conference on Electrical engineering and computer Science, ICECOS 2024

作者： Fuadi, Mukhlish Dharma Wibawa, Adhi Sumpeno, Surya Institut Teknologi Sepuluh Nopember Dept. of Electrical Engineering Surabaya Indonesia Institut Teknologi Sepuluh Nopember Dept. of Electrical Engineering Dept. of Medical Technology Surabaya Indonesia Institut Teknologi Sepuluh Nopember Dept. of Electrical Engineering Dept. of Computer Engineering Surabaya Indonesia

ISBN: (纸本)9798350368253

Question Answering (QA) is a technology that can help find answers to questions based on predefined information sources or context. However, QA sometimes still provides answers even though the question is irrelevant to the context. This inappropriate and irrelevant answer can certainly be misleading. Therefore, there is a need for QA that not only answers questions but can also recognize when the question is unanswerable. This study proposes a Text-To-Text Transfer Transformer (TS)-based QA model trained using a dataset that includes questions with and without answers. To evaluate the performance of this model on questions with answers, we compare it to a model that is only trained on questions with answers. In addition, to evaluate its performance on questions with answers and without answers, we compared it with previous research on the SQuAD-T test set and achieved an accuracy score of 77.24. The experimental results show that our proposed QA model is not only able to answer questions but also recognizes and handles questions without answers with better accuracy. © 2024 IEEE.

关键词： answer validation fine-Tuning question answering transformer

来源：评论

学校读者我要写书评

暂无评论

AI-Enhanced Virtual Trial Room using Augmented Reality 6

AI-Enhanced Virtual Trial Room using Augmented Reality

引用

6th International Conference on Mobile Computing and Sustainable informatics, ICMCSI 2025

作者： Mythili, N. Harini, M. Janalyn maroula, L. Dept. of Computer Science and Engineering St. Joseph’s Institute of Technology Old Mamallapuram Road 119 Chennai India

ISBN: (纸本)9798331522667

The Virtual Trial Room is an innovative augmented reality (AR) system that enables users to virtually try on apparel, enhancing the online shopping experience by simulating a physical fitting room. This technology leverages advanced computer vision, 3D modeling and machine learning to deliver an accurate and immersive experience. Data augmentation techniques are utilized to expand the training dataset, ensuring more precise clothing placement on different body types. The results demonstrate that our integrated approach, which combines AR and machine learning, significantly outperforms traditional virtual try-on systems, achieving enhanced user satisfaction, fit accuracy and personalized recommendations. The convolutional neural network (CNN) used in the system accurately detects body and facial landmarks, allowing for real-time garment fitting and recommendations. © 2025 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Vertical Federated Learning in Malware Detection for Smart Cities 9

Vertical Federated Learning in Malware Detection for Smart C...

引用

9th IEEE International Smart Cities Conference, ISC2 2023

作者： Serpanos, Dimitrios Xenos, Georgios Computer Technology Institute and Press Diophantus University of Patras Dept. of Electrical and Computer Engineering Patras Greece Computer Technology Institute and Press Diophantus University of Patras Campus Patras Greece

ISBN: (纸本)9798350397758

Malware detection is fundamental to smart city cyberphysical systems, considering their requirements for safety which is dependent on their security. Malware detection systems extract features from software samples through static and/or dynamic analysis and classify them as malware or benign, based on the features. Modern malware detection systems employ Deep Neural Networks (DNNs) whose accuracy increases as more data are analyzed and exploited. However, data sharing among organizations, even for malware, is limited due to privacy and intellectual property constraints. In this paper we investigate the effectiveness of vertical federated learning (VFL) for developing aggregated DNNs among autonomous interconnected organizations. We analyze a solution where multiple organizations use independent malware analysis platforms as part of their Security Operations Centers (SOCs) and train their own local DNN model on their own private data. Employing a DNN, we investigate VFL scalability in terms of the number of participants, considering accuracy and training execution time as the determining factors. We evaluate the approach using the EMBER benchmark dataset and demonstrate that the accuracy of a centralized model is achieved by VLF as well, independently of the number of clients, while training execution time scales linearly for the evaluated federation sizes and remains within acceptable bounds for adoption in practice. Furthermore, we evaluate the achieved accuracy per client (federation member) demonstrating that individual accuracy is significantly smaller per client. The significant improvement in accuracy that clients achieve by participating in the federation provides strong motivation for clients to participate in the federation, while FL provides them with the necessary privacy that is achieved by sharing features rather than samples. © 2023 IEEE.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Devise Road Sign Alert Detection for Vehicular Systems Using Fog Computing 3

Devise Road Sign Alert Detection for Vehicular Systems Using...

引用

3rd IEEE Asian Conference on Innovation in technology, ASIANCON 2023

作者： Jadhav, Saurabh Mahesh Abraham, Prof. Jibi Matange, Ashwini Coep Technological University Dept. of Computer Engineering and Information Technology MH Pune India

ISBN: (纸本)9798350302288

There are web mappings service applications like Google Maps, Apple Maps, and Waze that offer satellite imagery, street maps, real-time traffic conditions, departure arrival timings, and route planning for traveling by foot, car, bike, and public transportation. These applications enable users to view and hear traffic alerts, turn instructions, lane guidance, and available alternate route options. Road signs like speed breakers, sharp turns, accident-prone zone, steep slopes, etc. are placed beside or above streets to assist the drivers. These applications usually run on the display screen of the car or the commuter's mobile. The driver must keep an eye on the screen as well as on the road signs. At times the driver might miss the road sign due to high speed leading to a tense situation of varying severity. To overcome this problem, we propose an application in which the moving car is connected to a nearby Fog Node that stores road signs along with their proper location. Fog computing minimizes the amount of data transferred to the cloud, lowering bandwidth consumption and associated expenses.. Because the initial data processing occurs near the device, latency is reduced, and overall response time is improved. As the car moves closer to a road sign, The Fog Node evaluates the distance between the road sign and the car's geolocation. If the car is approaching a road sign, the Fog Node will generate an alert voice message. Fog Node being closer to the car will have minimum latency. Using Fog Computing we can send an early voice alert to a user within a certain time before he reaches the location. The solution proposed will provide timely alerts to users before reaching the road segment that requires the attention of the driver. © 2023 IEEE.

关键词： Fog

来源：评论

学校读者我要写书评

暂无评论

Securing Medical Records Using Private Blockchain technology 5

Securing Medical Records Using Private Blockchain Technology

引用

5th International Conference on Blockchain Computing and Applications, BCCA 2023

作者： Alothman, Basil Alenezi, Mariam Alhamdan, Danah Almutairat, Mariam Almousa, Zahraa Joumaa, Chibli Khan, Murad Kuwait College of Science and Technology Dept. of Computer Science and Engineering Kuwait-City Kuwait

ISBN: (纸本)9798350339239

The term 'electronic medical records' (EMRs) refers to information that is highly confidential and saved electronically. This information is associated with the diagnosis and treatment of a patient, and it must be routinely distributed among the concerned people such as doctors and nurses. The disclosure of private medical information or its alteration during an operation makes it extremely difficult for participants to exchange their medical files with one another. The most effective strategy for overcoming these obstacles is to implement an electronic medical records system that is based on blockchain technology. Within the scope of this study, we have addressed how the technology of blockchain might contribute to improved healthcare data management. Through the application of blockchain technology, the purpose of this research is to establish a connection between the health ministries and departments and both public and private hospitals in order to simplify the process of gaining access to one's medical records and transferring those records while maintaining patient confidentiality and improving safety. © 2023 IEEE.

关键词： Information management

来源：评论

学校读者我要写书评

暂无评论

Blockchain-Based Secure e-KYC Solution for Banking System 16

Blockchain-Based Secure e-KYC Solution for Banking System

引用

16th IEEE International Conference on Computational Intelligence and Communication Networks, CICN 2024

作者： Tabassum, Zoya Arsheen, Sobiya Ahmad, Khaleel Ozdemir, Gulcihan Maulana Azad National Urdu University Dept. of Computer Science and Information Technology Hyderabad India Informatics Institute Istanbul Technical University Istanbul Turkey

ISBN: (纸本)9798331505264

Blockchain technology offers a secured, digital, decentralized, ever-growing ledger that keeps records tamper-proof. The distributed ledger means that, even if one node is attacked or hacked, data may be accessed from other nodes, making the system more resistant to attacks and ensuring data availability. The hash function is used to strengthen the data integrity. E-KYC stands for electronic Know Your Customer, a way to verify a customer's identity. The banks frequently require the customer's identity documents for verification. Therefore, they spend an unwanted cost, effort, and abundant time during the KYC process. Sometimes, the information stored locally can be misleading. With unique features, blockchain-enabled e-KYC systems can protect sensitive data, reduce data breach risks, and ensure the integrity and confidentiality of e-KYC procedures. With its distributed ledger, blockchain technology stores e-KYC data across nodes and cryptographic techniques and can effectively increase e-KYC safety. Decentralization and immutability make blockchain technology a strong foundation for securing sensitive customer data. Strong authentication and public key cryptographic methods to protect e-KYC data. © 2024 IEEE.

关键词： Distributed ledger

来源：评论

学校读者我要写书评

暂无评论

Enhancing UAV capabilities for Object Picking: A flying Manipulator with Stiffness Control 8

Enhancing UAV capabilities for Object Picking: A flying Mani...

引用

8th International Conference on Control, Robotics and Cybernetics, CRC 2023

作者： Sewalt, Marinus Johannes Tim Woo, Kam The Hong Kong University of Science and Technology Dept. of Electronic and Computer Engineering Hong Kong

ISBN: (纸本)9798350330588

Unmanned aerial vehicles (UAVs) have gained widespread popularity in various applications. However, their physical interaction capabilities, especially object picking, remain a challenge. This paper presents the system overview of a novel flying manipulator, combining a quadrotor platform with a direct-drive gripper. The study examines the viability of high-speed scooping based on stiffness control with a flying manipulator. By leveraging stiffness control, the gripper offers a significant advantage over rigid grippers by avoiding large contact forces that may destabilize the flying platform. This enables efficient and reliable retrieval of flat objects from the ground without the need for overly cautious maneuvers. Successful results are demonstrated, showcasing the practicality and effectiveness of the gripper-drone integration for physical interactions. This development opens avenues for further advancements in UAV capabilities and expands their potential for real-world applications. © 2024 IEEE.

关键词： Manipulators

来源：评论

学校读者我要写书评

暂无评论

Multivariate Imputation by N Neighbour Mean and Chained Equation for Time Series Missing Data 2

Multivariate Imputation by N Neighbour Mean and Chained Equa...

引用

2nd IEEE International Conference on Industrial Electronics: Developments and Applications, ICIDeA 2023

作者： Pansari, Risham Kumar Gupta, Muktesh Wadhvani, Rajesh Maulana Azad National Institute of Technology Dept. of Computer Science and Engineering Bhopal India

ISBN: (纸本)9798350381979

Missing or incomplete data introduces numerous challenges during data analysis. If there exists any incomplete data then algorithms used for analysis can produce incorrect results. Data with missing or incomplete values may have a negative effect on computation. Thereby leading to an incorrect result. Certain algorithms are incapable of handling missing data properly, while other methods produce an efficient output to calculate the missing values. Because different machine learning algorithms become less effective when given incomplete data, it is extremely important to manage missing data properly. It could be possible that the data set might contain some missing values for various features due to the following reasons data were maintained in the file, data got corrupted, etc. This paper proposed a method that makes use of N neighbors of the data that is missing and makes use of the machine learning model to tune the computed value so as to impute a correct value for the missing data when the missing data is spread across different features. The proposed model makes use of the spatial characteristics of the time series data in order to predict missing data with the least error. Through experimentation, the proposed model demonstrated a high degree of accuracy in predicting missing values, even when the gaps spanned multiple feature columns. © 2023 IEEE.

关键词： Time series

来源：评论

学校读者我要写书评

暂无评论

Underwater Image Enhancement by Using a Colour-Correction and Dehazing Approach

引用

Procedia computer Science 2025年 259卷 818-825页

作者： Shubhankar Nalawade Mohit Dua Shubhankar Nalawade Dept. of Computer Engineering National Institute of Technology Kurukshetra India Mohit Dua Dept. of Computer Engineering National Institute of Technology Kurukshetra India

Capturing clear images underwater is difficult because light gets absorbed and scattered, resulting in degraded image quality with reduced visibility and colour inaccuracy. This research paper proposes a comprehensive approach that combines colour correction and dehazing techniques to enhance such deteriorated underwater images. In the colour correction step, principles of colour constancy are utilized to address colour distortions caused by underwater conditions. This correction significantly improves colour accuracy and enhances the visual fidelity of the image by first converting the RGB image into CIELab colour space and then performing white balancing in the CIELab colour space. This step successfully eliminates colour casts and restores the natural colours of underwater scenes, leading to visually pleasing and accurate representations. In the dehazing step, our method utilizes a dark channel prior-based approach to remove the haze by estimating the overall atmospheric light and the transmission map. The dehazing step significantly reduces haze, enhancing the visibility of submerged objects and fine details. Our technique has been evaluated subjectively and objectively, and the results have demonstrated its efficacy in reducing haze, correcting colour cast, and preserving the natural appearance of the image. The enhanced results consistently achieve great results for RMSE, PNSR, SSIM, and UCIQE, surpassing several existing underwater enhancement methods.

关键词： Underwater Image Enhancement Colour Correction Dark Channel Prior

来源：评论

学校读者我要写书评

暂无评论

Auto Disturbance Rejection Control Based on A Neural Network Predictive State Observer 15

Auto Disturbance Rejection Control Based on A Neural Network...

引用

15th International Conference on Intelligent Human-Machine Systems and Cybernetics, IHMSC 2023

作者： Du, Shengzhi Van Wyk, B.J. Nahri, N.F.S. Tu, Chunling Tshwane University of Technology Dept. Electrical Engineering Pretoria South Africa Tshwane University of Technology Dept. Computer Systems Engineering Pretoria South Africa

ISBN: (纸本)9798350326178

This paper addresses the auto disturbance rejection control (ADRC) for systems with time delay. ADRC is an effective control for most linear or nonlinear systems with noises and disturbances, because of its advantage on disturbance estimation and compensation. The extended state observer (ESO) is a critical component of ADRC. However, this observer is usually designed for systems without time delay. The time delay in a system could lead the ADRC unstable when the phase delay of the response cannot be ignored. This paper proposes a neural network based predictive state observer (NNPSO) to predict the response of systems with time delay. The simulation results validate the proposed method when systems with time delay are controlled. © 2023 IEEE.

关键词： Time delay

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：