检索结果-内蒙古大学图书馆

Advances in Computing, Communication and Networking (ICAC2N), International Conference on

作者： Palvi Sharma Rakesh Kumar Meenu Gupta Ahmed J. Obaid Dept. of Computer Science & Engineering Chandigarh University Punjab India Dept. of Computer Science & Engineering Chandigarh University Punjab Faculty of Computer Science and Mathematics University of Kufa Najaf Iraq College of Technical Engineering National University of Science and Technology Dhi Qar Iraq

ISBN: (数字)9798350356816

ISBN: (纸本)9798350356823

Accurate road extraction from Remote Sensing Imagery (RSI) is crucial for various applications, including urban planning, navigation, and disaster response. This study investigates automated road extraction in rural, semi-urban and urban regions using deep learning architectures, specifically UNet and Deep Residual U-Net (ResNet) models. We leverage High-Resolution Satellite Imagery (HRSI) to train and validate our models, focusing on the distinct characteristics and challenges presented by urban and rural environments. Our methodology involves comprehensive data preprocessing and augmentation techniques to enhance model robustness. We evaluate the performance of both architectures using metrics and achieved an accuracy of 90.29%, recall of 80.95%, precision of 74.89%, and F1- score of 77.80% for the Deep Residual U-Net. Experimental results demonstrates that while both models perform well, the Deep Residual U-Net exhibits superior accuracy and generalization, particularly in complex urban landscapes.

关键词： Training Accuracy Roads Computational modeling Urban planning computer architecture Feature extraction Satellite navigation systems Sensors Remote sensing

来源：评论

学校读者我要写书评

暂无评论

High-Performance Image Restoration with SpectraNet: A Lightweight CNN Architecture

High-Performance Image Restoration with SpectraNet: A Lightw...

引用

3rd International Conference on Machine Learning and Data engineering, ICMLDE 2024

作者： Patnaik, Nitesh Raj, Rishi Singh, Piyush Kumar Mahapatra, Shreekara Pani, Debashis Singh, Ravi Shankar Dept. of Computer Science and Engineering Kalinga Institute of Industrial Technology Bhubaneswar751024 India Dept. of Computer Science Central University of South Bihar Bihar 824236 India Dept. of Computer Science and Engineering Indian Institute of Technology BHU Varanasi221005 India

Deep learning-based image restoration faces significant challenges when deployed on resource-constrained platforms due to the computational demands and large number of parameters in existing models. The high computational load and extensive memory requirements make it difficult to implement these models on devices with limited processing power and storage capacity, such as mobile phones and embedded systems. Additionally, maintaining real-time performance while ensuring high-quality image restoration is a critical challenge, as traditional deep learning models often fail to meet the stringent latency and efficiency requirements of such platforms. This paper introduces a new, lightweight convolutional neural network (CNN) architecture tailored for efficient pixelation detection and restoration. Our approach combines a pre-trained MobileNetV2 with finetuning for pixelation detection, and SpectraNet, a architecture incorporating depth-wise separable convolutions for image restoration. The proposed architecture was trained on the HQ-50k dataset, with 70% of the data used for training and 30% for testing. It achieved a Peak Signal-to-Noise Ratio (PSNR) of 17-23, outperforming 10 state-of-the-art architectures in both efficiency and image quality. This design addresses the dual problems of high computational load and extensive memory usage, prioritizing real-time performance and resource efficiency. The proposed architecture is ideal for deployment on mobile and embedded devices since it maintains high accuracy while significantly reducing the number of trainable parameters. Our findings advance the feasibility of deploying advanced image restoration techniques in real-world, resource-constrained environments. © 2024 The Authors. Published by ELSEVIER B.V.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Blockchain-Based Resilient Pairing and Bonding of BLE Devices Using Deep Reinforcement Learning

引用

IEEE Transactions on Consumer Electronics 2024年

作者： Devi, Aguru Aswani Babu, Erukala Suresh Rathore, Rajkumar Singh Jhaveri, Rutvij H. Benedetto, Francesco SRM University AP Dept. of Computer Science and Engineering India National Institute of Technology Warangal Dept. of Computer Science and Engineering Telangana India Cardiff Metropolitan University Dept. of Computer Science Cardiff School of Technologies United Kingdom Pandit Deendayal Energy University Computer Science and Engineering School of Technology Gandhinagar India Roma Tre University Signal Processing for Telecommunications and Economics Department of Economics Rome Italy

In the day-to-day deployment of the Internet of Things (IoT), IPv6-over-Bluetooth Low Energy (BLE) devices became significant enablers due to their low power, low range, and effortless connectivity. To overcome the latency during BLE device migration, we have proposed a Blockchain-based resilient pairing and bonding of BLE devices using a lightweight authenticated encryption scheme. The bonding information is stored on the local ledger of the intermediate IPv6 over Low power Wireless Personal Area Networks (6LoWPAN) for BLE (6LoWPAN for BLE or 6LoWBLE) over gateway. Whenever the BLE device migrates from one gateway to another, the bonding information is transferred to the local ledger of the current gateway through the global ledger with the Proof-of-Voting consensus algorithm. The resilience of the proposed authentication scheme is analyzed using BAN logic. In addition, we proposed a Man-in-the-Middle (MITM) detection framework using Deep Reinforcement Learning (DRL) due to the escalation of MITM attacks on BLE devices. The DRL-based model is implemented in Python, and its performance is evaluated using the Kitsune Network Attack Dataset. The performance of the overall framework is tested on the Hyperledger Fabric, which results in low latency and a low average pairing time in comparison with existing frameworks. © 1975-2011 IEEE.

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Highly Efficient Switched Capacitor Converters for Rechargeable Hearing Aids 3

Highly Efficient Switched Capacitor Converters for Rechargea...

引用

3rd International Conference on Communication, Control, and Intelligent Systems, CCIS 2024

作者： Anandi, V. Muthuswamy, Ramesh Pranav, A. Ramaiah Institute of Technology Dept. of Electronics and Communication Bengaluru India School of Management Presidency University Bengaluru India Vit Dept. of Computer Science Engineering Chennai Campus Chennai India

ISBN: (纸本)9798331528201

The paper deals with switched capacitor-based designs for power-efficient rechargeable hearing aids. Switched-capacitor DC-DC power converters are a subset of DC-DC power converters that use only switches and capacitors to convert one voltage to another efficiently. By using switched-capacitor circuits, the aim is to increase the efficiency of hearing aids. To achieve long battery life, a high conversion efficiency is required. At the same time, the physical size should be low as it is desirable for the end-user to have small hearing aids. The voltage levels thus have to be stepped down from the voltage provided by a Li-ion (rechargeable) to levels closer to that of Zn-Air (non-rechargeable). © 2024 IEEE.

关键词： Capacitor bank

来源：评论

学校读者我要写书评

暂无评论

Adaptive Sliding Mode Control for Interconnected Power Systems Networks with Disturbances

Adaptive Sliding Mode Control for Interconnected Power Syste...

引用

2023 IEEE International Conference on Industrial Technology, ICIT 2023

作者： Islam, Shafiqul NCF-263 Xavier University of Louisiana Dept. of Dual-Degree Computer Science and Engineering New OrleansLA70125 United States

ISBN: (纸本)9798350336504

In this paper, we present adaptive sliding mode-based load frequency control for interconnected electric power systems networks with the presence of disturbances. The design comprises a state and sliding mode feedback vector with robust and adaptive control terms to learn and compensate for disturbances. The disturbances in power networks may appear from load variations, modelling errors and other external fault disturbances. The power networks may also face additional disturbance from the changes in the power system structure for bilateral contracts between distribution and transmission companies for the future open and liberal electricity trading policy. The design and analysis are developed by using the Lyapunov method. It is shown in our analysis shows that the states of the power networks are bounded and their bounds asymptotically converge to zero in the Lyapunov sense. Compared with existing results, the proposed method does not require the exact bounds of the disturbances. The design is simple and easy to implement as it does not require the exact model of the interconnected power systems networks. Evaluation results on three-area power systems with extreme conditions are presented to illustrate the effectiveness of the proposed method for real-time applications. © 2023 IEEE.

关键词： Electric power system interconnection

来源：评论

学校读者我要写书评

暂无评论

Evaluating Path Selection Strategies with Blockchain-based Routing in Multi-Domain SDNs

Evaluating Path Selection Strategies with Blockchain-based R...

引用

2022 International Balkan Conference on Communications and Networking, BalkanCom 2022

作者： Guler, Evrim Karakus, Murat Uludag, Suleyman Bartin University Dept. of Computer Engineering Bartin Turkey Bayburt University Dept. of Computer Technologies Bayburt Turkey U. of Michigan - Flint Dept. of Computer Science FlintMI United States

ISBN: (数字)9781665487641

ISBN: (纸本)9781665487641

Software-Defined Networking (SDN) has emerged as one of the most prolific inventions in the general area of computer communications over the past decade. Among many of the touted benefits, SDN architecture facilitated by OpenFlow has the potential to help network administrators to provide End-to-End (E2E) Quality of Service (QoS)-guaranteed paths for flows among networks while providing finer-granular flow management along with global network view. In our previous work, we have introduced QoSChain to amalgamate the advantages of blockchain and SDN for inter-network QoS provisioning to orchestrate an agile and software-driven traffic management framework in order to facilitate cross-network policy automation, assurance, E2E visibility, control and validation. In this study, we focus on the impact of the path selection strategies on the performance of the overall QoSChain performance. Specifically, we exploit three simple but effective path selection strategies, First Feasible Path Selection (FFPS), Random Feasible Path Selection (RFPS), and Minimum Hop Path Selection (MHPS), to evaluate their performance, impact, and overhead in terms of the flow setup time, message overhead, and network resource consumption metrics. Our preliminary simulations results show that the path selection plays an important role in the overall performance with hop count minimization delivering a superior performance offset by an initial higher setup times. © 2022 IEEE.

关键词： Quality of service

来源：评论

学校读者我要写书评

暂无评论

Recognition of Fingerprint Images using CNN for Cybercrime Detection System

Recognition of Fingerprint Images using CNN for Cybercrime D...

引用

Network, Multimedia and Information Technology (NMITCON) International Conference on

作者： T. Kavitha Bhimaraya Patil S. Saraswathi S. Rajarajeswari Anita Patil Dept. of Computer Engineering New Horizon College of engineering Bangalore India Dept. of Computer Science Engineering Sri Sivasubramaniya Nadar College of Engineering Chennai India School of computer science and engineering Vellore Institute of Technology Vellore India Dept. of Information Science City Engineering College Bengaluru India

ISBN: (数字)9798350372892

ISBN: (纸本)9798350372908

Fingerprint identification plays a crucial role in crime detection, but manual analysis is time-consuming and error-prone. This research aims to develop an automated system for accurate and efficient fingerprint matching using machine learning. The proposed approach involves image acquisition, preprocessing, feature extraction, and classification. Fingerprint images are enhanced, binarized, and segmented. Minutiae (ridge endings and bifurcations) are extracted as features. A convolutional neural network (CNN) is trained on a dataset of fingerprint images for classification. By comparing algorithms, Random Forest achieved over $\mathbf{8 1 \%}$ accuracy, outperforming Decision Trees (nearly 75% accuracy). The trained CNN model can accurately match fingerprints from crime scenes with potential suspects, providing reliable evidence. The developed system offers law enforcement agencies an automated and accurate fingerprint identification tool, improving crime-solving efficiency and reducing manual effort. The proposed approach showcases the potential of machine learning in forensic applications.

关键词： Accuracy Machine learning algorithms Image recognition Image matching Manuals Fingerprint recognition Feature extraction Convolutional neural networks Reliability Random forests

来源：评论

学校读者我要写书评

暂无评论

Relative Position Estimation Using Cross Filter for Rolling-Shutter Cameras

Relative Position Estimation Using Cross Filter for Rolling-...

引用

IEEE Conference on Global Communications (GLOBECOM)

作者： Daisuke Sekimoto Koji Kamakura Masayuki Kinoshita Takaya Yamazato Dept.of Computer Science Chiba Institute of Technology Narashino Japan Institute of Liberal Arts and Sciences Nagoya University Nagoya Japan

ISBN: (数字)9798350351255

ISBN: (纸本)9798350351262

We propose an estimation method of light emitting diode(LED) position with a cross filter (CF) for rolling-shutter (RS) camera. The CF is set in front of the camera lens so that incident light can be scattered into two straight lines extending ±π/4 from the horizontal row of the CMOS image sensor (IS). With the geometry of a right-angled isosceles triangle, our method makes it possible to estimate the coordinates of the LED in the IS plane when the scattered light is captured at two points in a scanline. Furthermore, our method makes it possible rows by rows when multiple LEDs are used or when they are moving. Without the CF, estimation is done only once per frame when the scan of the sequentially activated operation of the RS camera reaches a row that has just captured the whole image of the target LED, and the estimated coordinates are not updated in that frame. Experiments verify that our estimation method works with the estimation error less than 5pixels, and to extend the region of a frame in which estimated coordinates are output, even in a moving environment, while the conventional method (without CF) obtains the estimated coordinates only once when the scan reaches a row that captures entire LED.

关键词： Geometry Estimation error CMOS image sensors Light emitting diodes Cameras Global communication Lenses

来源：评论

学校读者我要写书评

暂无评论

Silent Expressions Unveiled: Deep Learning for British and American Sign Language Detection

引用

Procedia computer science 2024年 233卷 269-278页

作者： Sai Niketh Koyineni Gurram Kumar Sai Kalwa Anvesh T Anjali Dept. of Computer Science and Engineering Amrita School Of Computing Amrita Vishwa Vidyapeetham Amritapuri India

Sign language is an efficient means of communication for the hearing impaired. The main goal of this research is to recognize both American and British sign languages using computer vision algorithms and neural networks. Our work addresses the critical need for inclusive and accessible communication for the Deaf community. For the study, we compiled a large dataset of her ASL and BSL gestures. Preprocess the data to improve feature extraction and reduce noise. Our model architecture uses a convolutional neural network (CNN) to capture the temporal patterns of finger gestures. This allows you to use Mediapipe keypoint detection within your model. The end result is a comprehensive model that can accurately and quickly recognize and classify sign language gestures. We further improve the usability of the system by integrating keyword recognition into the model, allowing sign language sentences to be decoded and translated into text. Our test results demonstrate how well our deep learning strategy works to achieve high efficiency and accuracy in both ASL and BSL detection. This program helps bridge the gap between those who use sign language and those who do not. Our new approach not only accurately recognizes American and British sign language, but also seamlessly integrates real-time feedback to help users make immediate improvements to improve their sign language skills. We can do better. Moreover, our approach is flexible enough to be used in a variety of settings, ensuring its usefulness in educational environments and providing sign language learners with a fun and dynamic learning environment. Explore and exploit the full potential of Mediapipe Keypoint using CNNs and also leverage the efficiency of computer vision.

关键词： Sign language recognition BSL ASL deep learning neural networks computer vision gesture recognition assistive technology communication accessibility keyword detection Convolutional Neural Network(CNN)

来源：评论

学校读者我要写书评

暂无评论

Design and Analysis of a High-Gain Microstrip Patch Antenna Enhanced by Near-Zero Index Metamaterial Superstrate 8

Design and Analysis of a High-Gain Microstrip Patch Antenna ...

引用

8th IEEE International Conference on Information and Communication Technology, CICT 2024

作者： Kumar, Awanish Bhardwaj, Ajay Padhi, Jyotibhusan CEA-LETI University of Grenoble Alpes Grenoble France Srm University Ap Dept. of Computer Science and Engineering Amaravati India Srm University Ap Dept. of Ece Amaravati India

ISBN: (纸本)9798331531539

Seeing the rising applications of metamaterial in sensing and imaging, satellite communications, it becomes evident to design a high-gain microstrip patch antenna. To support these applications, this paper proposes a 7 x 7 array of planar novel metamaterial unit cells used as a superstrate to enhance the gain of microstrip patch antenna operating at 11.2 GHz. This proposed metamaterial structure yields a very low (near zero) value of effective refractive index at 11.2 GHz. Hence, the superstrate behaves as a near zero-indexed-medium (NZIM) around this frequency. NZIM superstrate are very popular because of their ability to focus the radiation and by utilizing this property, a significant gain enhancement has been achieved in the usage of patch antennas. Numerical simulations have been conducted using the CST Microwave studio, and obtained results corroborate that NZIM superstrate when suspended over a microstrip patch antennas significantly improves the gain around the value of 7.5 dB at 11.2 GHz, and efficiency is also improved. © 2024 IEEE.

关键词： Microstrip antennas

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：