检索结果-内蒙古大学图书馆

Fuzzy clustering for electric field characterization and its application to thunderstorm interpretability

Digital Communications and Networks 2025年第2期11卷 299-307页

作者： Xu Yang Hongyan Xing Xinyuan Ji Wei Xu Witold Pedrycz School of Electronics and Information Engineering Nanjing University of Information Science and TechnologyNanjing 210044China Department of Electrical and Computer Engineering University of AlbertaEdmonton AB T6G 2R3Canada Systems Research Institute Polish Academy of Sciences00-901 WarsawPoland Department of Computer Engineering Faculty of Engineering and Natural SciencesIstinye University34010 IstanbulTurkey

Changes in the Atmospheric Electric Field Signal(AEFS)are highly correlated with weather changes,especially with thunderstorm ***,little attention has been paid to the ambiguous weather information implicit in AEFS *** this paper,a Fuzzy C-Means(FCM)clustering method is used for the first time to develop an innovative approach to characterize the weather attributes carried by ***,a time series dataset is created in the time domain using AEFS *** AEFS-based weather is evaluated according to the time-series Membership Degree(MD)changes obtained by inputting this dataset into the ***,thunderstorm intensities are reflected by the change in distance from a thunderstorm cloud point charge to an AEF ***,a matching relationship is established between the normalized distance and the thunderstorm dominant MD in the space ***,the rationality and reliability of the proposed method are verified by combining radar charts and expert *** results confirm that this method accurately characterizes the weather attributes and changes in the AEFS,and a negative distance-MD correlation is obtained for the first *** detection of thunderstorm activity by AEF from the perspective of fuzzy set technology provides a meaningful guidance for interpretable thunderstorms.

关键词： Atmospheric electric field(AEF) Thunderstorm Fuzzy C-means(FCM) Attribute

来源：评论

学校读者我要写书评

暂无评论

Deploying On-Device AIGC Inference Services in 6G via Optimal MEC-Device Offloading

IEEE Networking Letters

引用

IEEE Networking Letters 2024年第4期6卷 232-236页

作者： Zhou, Changshi Liu, Weiqi Han, Tao Ansari, Nirwan New Jersey Institute of Technology Department of Electrical and Computer Engineering NewarkNJ07102 United States Auburn University at Montgomery Department of Computer Science and Computer Information Systems MontgomeryAL36117 United States

From AI-assisted art creation to large language model (LLM)-powered ChatGPT, AI-generated contents and services are becoming a transforming force. It calls for the telecom industry to embrace the prospects of AIGC services and face the unique challenges posed by incorporating generative model services into the AI-native 6G wireless network paradigm. We propose enabling AIGC inference services on mobile devices by optimizing MEC-device computing offloading, through which AIGC task latency is minimized by reinforcement learning based policy agent in a computing resource constrained and bandwidth limited wireless environment. Simulation results are presented to demonstrate the performance advantage. © 2019 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Empirical techniques for effort estimation in designing effective ML models

引用

International Journal of computers and Applications 2025年第2期47卷 226-234页

作者： Padmanabha Reddy, Y.C.A. Kosuru, Sai Sathwik Sirisalla, Nageswara Rao Vivekananda, G.N. Perala, Venkata Akanksha Department of Artificial Intelligence and Machine Learning Chaitanya Bharati Institute of Technology Telangana Hyderabad India Department of CSE B V Raju Institute of Technology Telangana Narsapur India Department of Computer Science and Engineering K.S.R.M. College of Engineering AP Kadapa India School of Computer Science Engineering and Information Systems Vellore Institute of Technology Vellore India

Machine Learning Research often involves the use of diverse libraries, modules, and pseudocodes for data processing, cleaning, filtering, pattern recognition, and computer intelligence. Quantization of Effort Required for the above cumulative processes is rarely discussed in the existing works and The time to reach the desired level of the model's functionality is essential to gauge the environment training for the model training and pre-deployment testing. In this study, we empirically defined the manual-cum-computational effort required for the model development in terms of 2-time factors: time-to-live (time until 1st code modification) and time-for-modification (time between 2 codebase changes) taken together in a mathematical model to compute Effort Factor(quantitative measure of determining effort needed to design ML algorithms). The study is novel in terms of how the effort required to create ML models is calculated with a particular focus on the manual effort direction. The results and the findings obtained can be used for determining the total time needed to synthesize ML models and frameworks in terms of code change cycles and implementation strategies marking a 25% performance increase in the current method above the Standard Pipeline. © 2025 Informa UK Limited, trading as Taylor & Francis Group.

关键词： Pipeline codes

来源：评论

学校读者我要写书评

暂无评论

Smart Business with Intelligent Business Cards (SBIBC)

Smart Business with Intelligent Business Cards (SBIBC)

引用

2024 International Conference on System, Computation, Automation and Networking, ICSCAN 2024

作者： Kirubakaran, N. Preethi, S.T. Parijatham, R. Pramithi, R. Chennai Institute of Technology Department of Computer Science and Business Systems Chennai India S.A. Engineering College Department of Computer Science and Engineering Chennai India Madha Institute Engineering and Technology Department of Computer Science and Engineering Chennai India

ISBN: (数字)9798331510022

ISBN: (纸本)9798331510022

The research emphasizes the creation of a powerful and efficient system for the automaticextraction of contact information from physical calling cards through computer vision and information extraction techniques. This is achieved by integrating Tesseract Optical Character Recognition (OCR) with the YOLO (You Only Look Once) object detection algorithm, which acts as a dynamic combination for text and relevant elements identification and digitization from calling cards. The developed system uses client-server architecture to enable communication and data processing seamlessly. The server handles computationally intensive tasks such as object detection, OCR processing, and data validation while the client side has an interactive user interface that enables easy uploading of images and retrieval of structured contact information. The optimizations include this: YOLO is customized for detecting particular card elements (name, phone number, email) and enhancing the capabilities of Tesseract OCR to improve effectiveness for text extraction from varied card formats or fonts in the system. It is designed to be scalable and efficient, processing large datasets without affecting accuracy. This approach not only eases the digitization of calling card details into databases but also goes a long way in contributing to the domains of computer vision, object detection, and mechanized information extraction. It extends the scope to broad applications in automated data entry, customer relationship management (CRM), and digital archival. It takes forward for AI-driven sustaining technologies the integration of these techniques into practical solutions by overcoming the challenges of data extraction from real-world images. © 2024 IEEE.

关键词： Optical character recognition

来源：评论

学校读者我要写书评

暂无评论

Leveraging Convolutional Neural Network and Explainable AI for Enhanced Pneumonia Detection 1

Leveraging Convolutional Neural Network and Explainable AI f...

引用

1st International Conference on Sustainable Energy Technologies and Computational Intelligence, SETCOM 2025

作者： Murugan, Ramalingam Patel, Harshita Vellore Institute of Technology School of Computer Science Engineering and Information Systems Tamil Nadu Vellore India

ISBN: (纸本)9798331520540

Advances in machine learning and computer vision have significantly improved the diagnostic capabilities of medical imaging. Convolutional Neural Networks (CNNs) have emerged as a crucial tool for image classification, particularly in detecting pneumonia from chest X-ray images. Despite their high accuracy, CNNs face challenges in clinical adoption due to their lack of interpretability. Medical practitioners require a clear understanding of the decision-making process to trust AI-generated diagnoses. To address the aforementioned challenge, we propose pneumonia classification model incorporating CNN with a combination of the model and Grad-CAM, for enhanced interpretability. Grad-CAM visually outlines critical regions, influencing the prediction, so the clinician can easily confirm AI-based diagnoses with a more confident and credible approach. It is exclusively for pediatric pneumonia, an accurate and explainable machine for early interventions that can also work in resource-scarce regions. Our research uses a set of 5,863 pediatric chest X-ray images, divided into Pneumonia and Normal classes. The dataset was acquired from Guangzhou Women and Children's Medical Center and divided into three subsets: training, validation, and test. By improving the accuracy and transparency of the diagnosis, this model can help bridge the gap between AI and the real world. This can help facilitate quicker and more reliable pneumonia diagnostics. © 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

Enhancing Skin Disease Diagnosis: Light GBM-DMS Algorithm for Accurate Image Classification

引用

Cognitive Computation 2025年第3期17卷 1-18页

作者： Kalpana, B. Senthilselvi, A. Muruganandam, S. Kumar, S. Vinodh Department of Information Technology RMD Engineering College Tamil Nadu Kavaraipettai India Department of Computer Science and Engineering SRM Institute of Science and Technology Ramapuram Tamil Nadu Chennai India Department of Computer Science and Business Systems Panimalar Engineering College Poonamallee Chennai Tamil Nadu India Department of Computer Science and Engineering Rajalakshmi Engineering College Tamil Nadu Thandalam Chennai India

Globally, skin diseases are emerging as the most common health problem. It initiates depressive disorder, and it also causes physical health distress. It rarely led to skin cancer in extreme cases. Diagnosing skin disease is subjective and time-consuming when it is performed by health experts manually. Hence, automatic skin disease prediction is required by both dermatologists and patients to enhance and accelerate treatment plans. Even though many machine-learning techniques are available to detect skin diseases, the timely prediction and accurate classification of skin lesions are still in demand. Hence, the Discrete Mycorrhizal Search Algorithm is used to optimize the novel LightGBM to meet this objective. Firstly, intrusions like noise elimination and hair removal are performed and then the image resizing takes place by utilizing the preprocessing pipelines. Subsequent to data pre-processing, the prime features are extracted using the convolutional self-attention based L2 regularization, thus minimizing the data complexity. At last, the extracted features are shifted to the final categorization phase which is done by the novel approach for effective classification of skin diseases. Two skin disease classification images, including datasets, the Skin Lesions Dermatoscopic Image dataset and Skin Cancer MNIST HAM10000, are selected to validate the classification performance. Comprehensive analyses are performed with significant measures that showed the proposed model attained better performances of 98.6%, 0.95, and 10.43 s from accuracy, intersection of union, and execution time, respectively. The stability and robustness under different environmental conditions for clinical applications in skin disease detection and classification tasks are analyzed for making decisions in real-time scenarios. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2025.

关键词： Dermatitis

来源：评论

学校读者我要写书评

暂无评论

Gaussian Blur Masked ResNet2.0 Architecture for Diabetic Retinopathy Detection

引用

computers, Materials & Continua 2023年第4期75卷 927-942页

作者： Swagata Boruah Archit Dehloo Prajul Gupta Manas Ranjan Prusty A.Balasundaram School of Electronics Engineering Vellore Institute of TechnologyChennai600127India School of Computer Science and Engineering Vellore Institute of TechnologyChennai600127India Centre for Cyber-Physical Systems Vellore Institute of TechnologyChennai600127India

Diabetic Retinopathy (DR) is a serious hazard that can result inirreversible blindness if not addressed in a timely manner. Hence, numeroustechniques have been proposed for the accurate and timely detection ofthis disease. Out of these, Deep Learning (DL) and computer Vision (CV)methods for multiclass categorization of color fundus images diagnosed withDiabetic Retinopathy have sparked considerable attention. In this paper,we attempt to develop an extended ResNet152V2 architecture-based DeepLearning model, named ResNet2.0 to aid the timely detection of DR. TheAPTOS-2019 datasetwas used to train the model. This consists of 3662 fundusimages belonging to five different stages of DR: no DR (Class 0), mild DR(Class 1), moderate DR (Class 2), severe DR (Class 3), and proliferativeDR (Class 4). The model was gauged based on ability to detect stage-wiseDR. The images were pre-processed using negative and positive weightedGaussian-based masks as feature engineering to further enhance the qualityof the fundus images by removing the noise and normalizing the images. Upsamplingand data augmentation methods were used to address the skewnessof the original dataset. The proposed model achieved an overall accuracyof 91% and an area under the receiver-operating characteristic curve (AUC)score of 95.1%, outperforming existing Deep Learning models by around 10%.Furthermore, the class-wise F1 score for No DR was 92%, Mild DR was 82%,Moderate DR was 66%, Severe was DR 89% and Proliferative DR was 80%.

关键词： Diabetic retinopathy deep learning transfer learning image processing image classification

来源：评论

学校读者我要写书评

暂无评论

Collision Avoidance System using YOLO-based Object Detection and Distance Estimation 3

Collision Avoidance System using YOLO-based Object Detection...

引用

3rd International Conference on Intelligent Data Communication Technologies and Internet of Things, IDCIoT 2025

作者： Raj, Shiv Akash Dhinakar Singh, Ashita Shekhawat, Tarakshraj Singh Manav Muthanna, M. Sivasankaran, K. School of Electronics Engineering Vellore Institute of Technology Vellore India School of Computer Science Engineering and Information Systems Vellore Institute of Technology Vellore India

ISBN: (纸本)9798331527549

Road accidents are a significant global issue, with human error causing a substantial number of collisions. This emphasizes the need for advanced collision avoidance systems to improve road safety. In response, this study integrates object detection and distance estimation techniques to enhance vehicle safety. Utilizing recent iterations of the YOLO architecture—YOLOv6n, YOLOv8n, YOLOv9t, and YOLOv10n—the models were trained and evaluated on the KITTI dataset, a benchmark suite for driving applications. This research study provides a comparative analysis of these models, focusing on performance metrics such as precision, mean average precision, recall, and F1 score. A novel single-camera distance estimation method was also implemented to assess object proximity accurately. Among the evaluated models, YOLOv8n achieved the best results, with a F1 score of 0.74 and a mAP@0.5 of 0.782, demonstrating superior detection accuracy and reliability. Its swift convergence and low training losses make it well-suited for real-world use cases. While YOLOv9t and YOLOv10n also performed well, they were slightly less effective. The system enhances its capabilities by accurately estimating distances, which improves collision avoidance and road safety applications. Overall, YOLOv8n is the most efficient and reliable option for developing advanced safety features in vehicle systems. © 2025 IEEE.

关键词： Vehicle safety

来源：评论

学校读者我要写书评

暂无评论

Point Cloud Classification Using Content-Based Transformer via Clustering in Feature Space

引用

IEEE/CAA Journal of Automatica Sinica 2024年第1期11卷 231-239页

作者： Yahui Liu Bin Tian Yisheng Lv Lingxi Li Fei-Yue Wang the State Key Laboratory for Management and Control of Complex Systems Institute of AutomationChinese Academy of SciencesBeijing 100190 the School of Artificial Intelligence University of Chinese Academy of SciencesBeijing 100190 IEEE the Transportation and Autonomous Systems Institute(TASI)and the Department of Electrical and Computer Engineering Purdue School of Engineering and TechnologyIndiana University-Purdue University Indianapolis(IUPUI)Indianapolis 46202 USA

Recently, there have been some attempts of Transformer in 3D point cloud classification. In order to reduce computations, most existing methods focus on local spatial attention,but ignore their content and fail to establish relationships between distant but relevant points. To overcome the limitation of local spatial attention, we propose a point content-based Transformer architecture, called PointConT for short. It exploits the locality of points in the feature space(content-based), which clusters the sampled points with similar features into the same class and computes the self-attention within each class, thus enabling an effective trade-off between capturing long-range dependencies and computational complexity. We further introduce an inception feature aggregator for point cloud classification, which uses parallel structures to aggregate high-frequency and low-frequency information in each branch separately. Extensive experiments show that our PointConT model achieves a remarkable performance on point cloud shape classification. Especially, our method exhibits 90.3% Top-1 accuracy on the hardest setting of ScanObjectN N. Source code of this paper is available at https://***/yahuiliu99/PointC onT.

关键词： Content-based Transformer deep learning feature aggregator local attention point cloud classification

来源：评论

学校读者我要写书评

暂无评论

Advancing Plant Disease Prediction Accuracy with ViT Based Deep Learning Model

Advancing Plant Disease Prediction Accuracy with ViT Based D...

引用

2024 IEEE International Conference on Computing, Semiconductor, Mechatronics, Intelligent systems and Communications, COSMIC 2024

作者： Sam, Felicita Kavitha, B.R. School of Computer Science Engineering and Information Systems Vellore Institute of Technology TamilNadu Vellore India

ISBN: (纸本)9798331517892

Agriculture is the key source for many people's livelihoods and an important contributor to a country's economy. There is a huge demand to streamline the process in the field of agriculture by integrating Computing into the traditional way of farming. Computing plays a key role in handling this sector with many newer revolutions that would automate the various processes involved in overall farming. Particularly, machine and deep learning based models have shown remarkable results in smart farming. Rice crop is considered in this work which is the major crop in India. It is necessary to meet the demands of the growing population in terms of sufficient production of this crop. One of the key issues in this crop is the plant diseases that impact the overall crop yield production considerably. Among the various diseases, the most commonly occurring disease is White stem Borer. Diagnosing the disease at an early stage could treat the problem at the earliest to ensure the appropriate amount of yield production. In this paper, a Vision transformer (ViT) based deep learning model is proposed to predict the white stem borer disease on the PlantDoc dataset and the results obtained are compared with the other promising CNN based pre trained models like VGG19, InceptionV3, Resnet50. The experimental output shows that Vision transformer based model outperforms the other proven models with an accuracy of 96%. © 2024 IEEE.

关键词： Plant diseases

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：