检索结果-内蒙古大学图书馆

Informatica (Slovenia) 2024年第6期48卷 185-198页

作者： Bajić, Filip Habijan, Marija Nenadić, Krešimir University Computing Centre University of Zagreb Zagreb10000 Croatia Faculty of Electrical Engineering Computer Science and Information Technology Osijek Osijek31000 Croatia

Data’s role is pivotal in the era of internet technologies, but unstructured data poses comprehension challenges. Data visualizations like charts have emerged as crucial tools for condensing complex information. Classifying charts and applying various processing techniques are vital to interpreting visual data. Traditional chart image classification methods rely on predefined rules and have limited accuracy. The advent of support vector machines (SVMs) and convolutional neural networks (CNNs) significantly improved the accuracy of these methods. This research evaluates our previously introduced Shallow convolutional neural network (SCNN) architecture for chart image classification, comprising four convolutional layers, two max-pooling layers, and one fully-connected layer. The network achieves state-of-the-art results, requiring smaller datasets and reduced computational resources. When two networks are combined into Siamese SCNN (SSCNN), emphasizing generalization, it achieves high accuracy with small datasets and excels in open-set classification. The evaluation process encompasses the utilization of six publicly available datasets. © 2024 Slovene Society Informatika. All rights reserved.

关键词： Data visualization

来源：评论

学校读者我要写书评

暂无评论

Life and death of a data set: a forensic investigation

引用

Spectroscopy Europe 2023年第2期35卷 32-35页

作者： Davies, Antony N. SERC Sustainable Environment Research Centre Faculty of Computing Engineering and Science University of South Wales United Kingdom

Davies talks about spectroscopic data set. As originally delivered the spectrometer was capable of generating some of the best data sets people had ever been able to measure. Excellent signal-to-noise ratios and extremely stable calibrations. The associated computer hardware was somewhat behind the state-of-the-art, but this was quite normal due to the long development times for the instrument hardware. And so, in the hands of many expert, and some less expert, scientists, this wonderful spectrometer gave birth to many spectroscopic data sets--reinforcing theories and dispelling some myths. Over its lifespan serving as an excellent measurement platform allowing many modifications to the original basic equipment. So, while our data was young there were no problems. All the measurement parameters were stored with the data set so we could check the instrument had been set up correctly.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Advanced Pedestrian Detection in Low Visibility Scenarios Using UltraLytics YOLOv8 and Kalman Filtering 5

Advanced Pedestrian Detection in Low Visibility Scenarios Us...

引用

5th International Conference on Sustainable Communication Networks and Application, ICSCNA 2024

作者： Pattem, Sanjay Kumar Reddy Rajagopal, Manjula Nirujogi, Nandan SRM Institute of Science and Technology Computing Technologies Chennai India SRM Institute of Science and Technology Computing Technologies Faculty of Engineering and Technology Chennai India

ISBN: (纸本)9798331530013

Nighttime detection of pedestrians brings important problems such as low light conditions, limitations of sensors, and environmental noise. To tackle this problem, in this study, we propose a novel visual-radar data fusion to enhance a pedestrian detection system using the UltraLytics YOLOv8 deep learning model together with Kalman Filtering technique. The system is based on a Low-Light Vision Infrared Paired (LLVIP) dataset that includes infrared vision data to attenuate the performance decrement of a pedestrian detector in the dark. The combination of infrared cameras and millimeter-wave radar achieves a detection rate of 97.6%, in which UltraLytics YOLOv8 exhibits excellent object recognition performance by generating accurate spatial & category information from the infrared image;complementary distance and velocity data are obtained by millimeter-wave radar. It brings Kalman Filtering to track the radar-detected pedestrian targets, making the system more robust by predicting target and noise reduction. It was capable of fusing infrared and radar data at decision-level to make the system provide very precise position accurately in even nighttime situations. The experimental results show that the proposed model has improved capability of detection accuracy, robustness, and response time in comparison with the single-sensor system, which is suitable for practical autonomous driving scenarios. © 2024 IEEE.

关键词： Kalman filters

来源：评论

学校读者我要写书评

暂无评论

Crimes associated with geological materials: mining, minerals, metals and fossils

引用

Geology Today 2024年第4期40卷 159-164页

作者： Donnelly, Laurance John Pirrie, Duncan AHK International Limited and IUGS Initiative on Forensic Geology Faculty of Computing Engineering and Science University of South Wales

Civilization has always relied on geological materials and it would not exist as we know it without the use of minerals. For the foreseeable future, minerals will remain fundamentally important commodities. As technology changes, so does the demand for different mineral commodities. For example, as we enter the era of the battery revolution, expedited by vehicle electrification and a reduction in the burning of traditional fossil fuels (coal, oil and gas), demand for minerals used in battery technologies is predicted to increase significantly. In 2022 alone, the global mining industry had a value of over 2000 billion US dollars. Where such huge amounts of money are involved, it is unsurprising that criminal activity follows. Criminal activity in the mining, minerals and metals industries includes illegal mining and smuggling, theft (including adulteration and substitution), illegal trade in conflict minerals, fakes, fraud and environmental crime (e.g. pollution, contamination and degradation of the land, air and water). In this article, we provide an overview of the types of criminal activities associated with geological materials and the potential role of forensic geologists in mitigating the associated risks. © 2024 The Author(s). Geology Today published by The Geological Society of London and The Geologists’ Association and John Wiley & Sons Ltd.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Hybrid Deep Learning Based Attack Detection for Imbalanced Data Classification

引用

Intelligent Automation & Soft computing 2023年第1期35卷 297-320页

作者： Rasha Almarshdi Laila Nassef Etimad Fadel Nahed Alowidi Department of Computer Science Faculty of Computing and Information TechnologyKing Abdulaziz UniversityJeddahSaudi Arabia Department of Computer Science Faculty of Computer Science and EngineeringUniversity of HailHailSaudi Arabia

Internet of Things(IoT)is the most widespread and fastest growing technology *** to the increasing of IoT devices connected to the Internet,the IoT is the most technology under security *** IoT devices are not designed with security because they are resource constrained ***,having an accurate IoT security system to detect security attacks is *** Detection Systems(IDSs)using machine learning and deep learning techniques can detect security attacks *** paper develops an IDS architecture based on Convolutional Neural Network(CNN)and Long Short-Term Memory(LSTM)deep learning *** implement our model on the UNSW-NB15 dataset which is a new network intrusion dataset that cate-gorizes the network traffic into normal and attacks *** this work,interpolation data preprocessing is used to compute the missing ***,the imbalanced data problem is solved using a synthetic data generation *** experiments have been implemented to compare the performance results of the proposed model(CNN+LSTM)with a basic model(CNN only)using both balanced and imbalanced ***,with some state-of-the-art machine learning classifiers(Decision Tree(DT)and Random Forest(RF))using both balanced and imbalanced *** results proved the impact of the balancing *** proposed hybrid model with the balance technique can classify the traffic into normal class and attack class with reasonable accuracy(92.10%)compared with the basic CNN model(89.90%)and the machine learning(DT 88.57%and RF 90.85%)***,comparing the proposed model results with the most related works shows that the proposed model gives good results compared with the related works that used the balance techniques.

关键词： IoT IDS deep learning machine learning CNN LSTM

来源：评论

学校读者我要写书评

暂无评论

An ensemble deep learning model for cyber threat hunting in industrial internet of things

引用

Digital Communications and Networks 2023年第1期9卷 101-110页

作者： Abbas Yazdinejad Mostafa Kazemi Reza M.Parizi Ali Dehghantanha Hadis Karimipour Cyber Science Lab School of Computer ScienceUniversity of GuelphOntarioCanada Department of Electrical Engineering Faculty of EngineeringShahed UniversityTehranIran College of Computing and Software Engineering Kennesaw State UniversityGAUSA School of Engineering University of GuelphOntarioCanada

By the emergence of the fourth industrial revolution,interconnected devices and sensors generate large-scale,dynamic,and inharmonious data in Industrial Internet of Things(IIoT)*** vast heterogeneous data increase the challenges of security risks and data analysis *** IIoT grows,cyber-attacks become more diverse and complex,making existing anomaly detection models less effective to *** this paper,an ensemble deep learning model that uses the benefits of the Long Short-Term Memory(LSTM)and the AutoEncoder(AE)architecture to identify out-of-norm activities for cyber threat hunting in IIoT is *** this model,the LSTM is applied to create a model on normal time series of data(past and present data)to learn normal data patterns and the important features of data are identified by AE to reduce data *** addition,the imbalanced nature of IIoT datasets has not been considered in most of the previous literature,affecting low accuracy and *** solve this problem,the proposed model extracts new balanced data from the imbalanced datasets,and these new balanced data are fed into the deep LSTM AE anomaly detection *** this paper,the proposed model is evaluated on two real IIoT datasets-Gas Pipeline(GP)and Secure Water Treatment(SWaT)that are imbalanced and consist of long-term and short-term dependency on *** results are compared with conventional machine learning classifiers,Random Forest(RF),Multi-Layer Perceptron(MLP),Decision Tree(DT),and Super Vector Machines(SVM),in which higher performance in terms of accuracy is obtained,99.3%and 99.7%based on GP and SWaT datasets,***,the proposed ensemble model is compared with advanced related models,including Stacked Auto-Encoders(SAE),Naive Bayes(NB),Projective Adaptive Resonance Theory(PART),Convolutional Auto-Encoder(C-AE),and Package Signatures(PS)based LSTM(PS-LSTM)model.

关键词： Internet of things IIoT Anomaly detection Ensemble deep learning Neural networks LSTM

来源：评论

学校读者我要写书评

暂无评论

Image Caption Generation with Python: A Deep Learning-Based Approach 5

Image Caption Generation with Python: A Deep Learning-Based ...

引用

5th International Conference on IoT Based Control Networks and Intelligent Systems, ICICNIS 2024

作者： Krishna, Siram Chaitanya Reddy, Painti Nagi Kirubanantham, P. School of Computing SRM Institute of Science and Technology Kattankulathur Department of Computing Technologies Chennai India School of Computing College of Engineering and Technology SRM Institute of Science and Technology Kattankulathur Faculty of Engineering and Technology Department of Computing Technologies Chennai India

ISBN: (纸本)9798331518097

Today, the day-to-day generation of such a large amount of content on social media and other websites makes it impossible to maintain each image manually, so the need for automatically identifying sensitive images and filtering them out becomes inevitable. This project proposes a sensitive image classification application using automatic captions in identifying and filtering inappropriate content. Thus, the information extracted from images by ResNet50 in the caption generation process is allowed to integrate such features with a standard model through Senti-Long Short-Term Memory networks (SLSTM) for better captioning. Deep learning models bring about drastically improved quality of captions along with contextual understanding. The captions are tested rigorously with training loss metrics, BLEU scores, and sentiment analysis through the VADER Sentiment Analyzer. This system efficiently categorizes images involving violence, negativity, or dangerous situations into the class of negative. A Flask-based web application was developed for implementing a scalable and user-friendly interface, ensuring efficient content moderation. This project has illustrated the possibility of integrating deep learning techniques with natural language processing to yield beneficial content moderation, further ensuring safe and controlled online environments. © 2024 IEEE.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

A review of tracking concept drift detection in machine learning 4th

A review of tracking concept drift detection in machine lear...

引用

4th Annual International Conference on Data science, Machine Learning and Blockchain Technology, AICDMB 2023

作者： Abdu, Nail Adeeb Ali Basulaim, Khaled Omer Information Technology Department Faculty of Engineering & Computing University of Science and Technology Aden Yemen Information Technology Engineering Department Faculty of Engineering University of Aden

ISBN: (纸本)9781032426853

The availability of time series streaming data has increased dramatically in recent years. Since the last decade, there has been a growing interest in learning from realtime data. While extracting significant information from data streams, online learning faces changes in data distribution. Concept drift refers to hidden data contexts that learning systems are unaware of it. Using previous training instances from the data stream, the classifier categorizes new instances. Hence in this regard, it is obvious to have considerable methods to discover the similarity or diversity status of the present streaming data that compared to the buffered data. This issue is defined and explored in contemporary literature as concept drift. This paper was initiated to identify the constraints of the contemporary methods devised to handle the concept drift in data streams in relation to supervised learning, we also review and categories the concept drift detectors with their key points. Eventually, the article presented observations can provide decision boundary detection at improved levels and discussing open research challenges and possible new research direction. © 2024 The Author(s).

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Liberated Arabic Handwritten Text Recognition using Convolutional Recurrent Neural Networks

Iraqi Journal for Computer Science and Mathematics

引用

Iraqi Journal for Computer science and Mathematics 2025年第2期6卷 72-81页

作者： AlRababah, Ahmad AbdulQadir Aljahdali, Mohammed Khalid Al jahdali, Abdulrahim Abdulhamid AlGhanmi, Mohammed Saleh Al Barazanchi, Israa Ibraheem Department of Computer Science Faculty of Computing and Information Technology in Rabigh King Abdulaziz University Saudi Arabia College of Engineering University of Warith Al-Anbiyaa Karbala Iraq

Arabic script is exhibited in a cursive style, which is a departure from the norm in many common languages, and the shapes of letters are contingent on their positions within words. The form of the first letter is influenced by the subsequent letter, middle letters are shaped by both preceding and succeeding letters, and the shape of the final letter is determined by the preceding letter. Additionally, certain letters are found to have strikingly similar shapes, making Arabic text recognition a formidable challenge in computer vision. The challenge of detecting and recognizing Arabic handwritten text is addressed in this paper by proposing a novel system that integrates two Deep Learning models without the constraints of a predefined dictionary or language model. Object Detection techniques are employed in the first model to accurately identify lines, words, and punctuation marks, providing a robust foundation for text recognition. In the second model, a powerful recognition system is developed, trained on the IFN/ENIT dataset to predict character sequences from handwritten text images. The architecture is comprised of a convolutional neural network (CNN) with residual connections, followed by a Bi-directional Long Short-Term Memory (BLSTM) layer, and culminates in a fully connected layer. State-of-the-art results in unconstrained Arabic text recognition tasks are achieved by this approach. © 2025 The Author(s).

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Vehicle Detection in Challenging Scenes Using CenterNet Based Approach

引用

Computers, Materials & Continua 2023年第2期74卷 3647-3661页

作者： Ayesha Muhammad Javed Iqbal Iftikhar Ahmad Madini OAlassafi Ahmed SAlfakeeh Ahmed Alhomoud Department of Computer Science University of Engineering and Technology Taxila47050Pakistan Faculty of Computing and Information Technology King Abdulaziz University Jeddah21589Saudi Arabia Faculty of Computing and Information Technology Northern Border UniversityRafhaSaudi Arabia

Contemporarily numerous analysts labored in the field of Vehicle detection which improves Intelligent Transport System(ITS)and reduces road *** major obstacles in automatic detection of tiny vehicles are due to occlusion,environmental conditions,illumination,view angles and variation in size of *** research centers on tiny and partially occluded vehicle detection and identification in challenging scene specifically in crowed *** this paper we present comprehensive methodology of tiny vehicle detection using Deep Neural Networks(DNN)namely *** DNN disregards objects that are small in size 5 pixels and more false positives likely to happen in crowded *** there are two categories of deep learning models single-step and two-step.A single forward pass model is the one in which detection is performed directly to possible location over dense sampling,wherein two-step models incorporated by Region proposals followed by object *** in this research scrutinize one-step State of the art(SOTA)model CenteNet as proposed recently with three different feature extractor ResNet-50,HourGlass-104 and ResNet-101 one by *** train our model on challenging KITTI dataset which outperforms in comparison with SOTA single-step technique MSSD300∗which depicts performance improvement by 20.2%mAPandSMOKEby with 13.2%mAP *** of CenterNet can be justified through the huge improved *** performance of our model is evaluated on KITTI(Karlsruhe Institute of Technology and Toyota Technological Institute)benchmark dataset with different backbones such as ResNet-50 gives 62.3%mAP ResNet-10182.5%mAP,last but not the least HourGlass-104 outperforms with 98.2%mAP CenterNet-HourGlass-104 achieved high mAP among above mentioned feature *** also compare our model with other SOTA techniques.

关键词： CenteNet SOTA object detection deep learning MSSD SMOKE DNN KITTI

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：