检索结果-内蒙古大学图书馆

1st International Conference on Software, Systems and Information technology, SSITCON 2024

作者： Liu, Yutong Heilongiang University of Science and Technology School of Computer and Information Engineering Harbin China

ISBN: (纸本)9798350352931

This present work describes a framework for the analysis of score dynamics and energy fluctuations in competitive environments, aimed at forecasting match outcomes and providing strategic decisions useful to coaching and players' performance. Potential energy was defined, seven key measures were identified related to Kendall's tau, shown as heat maps to outline the underlying patterns. Using the CV method and the AHP composite weighting, overall scores can determine the overall possible energy level of players. The line graph should be drawn based on the latter in order to assess volatility throughout the course of the match. Then, it carries out a derivation method for potential energy series, followed by extracting the most critical turning points in energy dynamics through the PCA-HMM-BN algorithm. Feature engineering with lag features and effect coding improves predictability. This graph presents a comparison among different machine learning models, namely BP Neural Network, XGBoost, and SVM, by showing the high performance obtained by an SVM model tuned with the ASAA algorithm, with an R-squared score of 0.99967. © 2024 IEEE.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Lithium battery SOH and RUL prediction based on time-frequency multi-feature extraction and fusion networks 5

Lithium battery SOH and RUL prediction based on time-frequen...

引用

5th International Conference on computer Vision, Image and Deep Learning, CVIDL 2024

作者： Wang, Xiaohua Dai, Ke Hu, Min Ni, Nanbing Hefei University of Technology School of Computer Science and Information Engineering Hefei China

ISBN: (纸本)9798350373820

Accurate prediction of battery state of health (SOH) and remaining useful life (RUL) is crucial for reducing the risk of lithium battery failure and intelligent management of energy storage power plants. Currently, most of the existing research methods only take the features between charge/discharge cycles as model inputs, ignoring the interconnectivity of feature data within the battery cycle. A model based on time-frequency multi-feature extraction and fusion (TFMEFNet) is proposed to improve the accuracy of lithium-ion battery SOH and RUL prediction. The time-frequency multi-feature extraction module is used to extract features of time and frequency domains respectively from the data within the discharge cycle, and then the global, time-domain and frequency-domain features are fused by the feature fusion module. Long-term dependencies in continuous data then captured by a TCN-GRU (Temporal Convolutional Network-Gated Recurrent Unit) module. A Temporal Attention module is used to better represent the decaying trend of the capacity series, weighted by attention, resulting in more accurate predictions. Extensive experiments are conducted on NASA datasets and compared with existing methods. The experimental results show that the proposed model can predict the SOH and RUL of Li-ion batteries more accurately. The root mean square error (RMSE) on the NASA dataset is less than 0.786%. © 2024 IEEE.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Dynamic strategy to use optimum memory space in real-time video surveillance

引用

Journal of Ambient Intelligence and Humanized Computing 2023年第3期14卷 2771-2784页

作者： Biswas, Tamal Bhattacharya, Diptendu Mandal, Gouranga Department of Computer Science and Engineering National Institute of Technology Agartala Tripura Agartala India School of Computer Science and Engineering VIT-AP University Andhra Pradesh Amaravati India

In present days real-time video surveillance is a very essential aspect of establishing safety, observing traffic, and detecting violence and crimes. Intelligent video surveillance (IVS) is one of the most acknowledged frameworks in a security application. But, video surveillance systems consume huge memory space to store the recorded video, especially in the case of high-resolution recording. However, if we use low-resolution video recording then the video quality will be reduced and the objects/violence cannot be identified clearly. To overcome the stated limitation, this article introduced a smart video surveillance system with a dynamic strategy to use optimum memory space with the best surveillance objective. The proposed system can dynamically record video with high-resolution during suspicious movement of objects/violence as well as low-resolution video at other times. The system is based on the detection of suspicious movement of objects in consecutive frames, saving of suspicious frames with high resolution, and neglecting of less important frames. Finally, the Contrast Limited Adaptive Histogram Equalization (CLAHE) based on Color Channel is applied to all the suspicious frames as a post-processing step for better contract adjustment and to provide a clear view of suspicious objects in frames. The quality of enhanced frames are assessed by two no-reference methods, i.e. NIQMC and BIQME. Multiple visual quality parameters like sharpness, contrast, colorfulness, brightness, and naturalness of frames, etc. are used for assessment Several real-time experiments reveal that the proposed system can detect suspicious objects with 97.65% accuracy. Besides, it consumes 75% less memory space in real-time surveillance. © 2023, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.

关键词： Video recording

来源：评论

学校读者我要写书评

暂无评论

Digital twin empowered lightweight and efficient blockchain for dynamic internet of vehicles

引用

Digital Communications and Networks 2024年第6期10卷 1698-1707页

作者： Haoye Chai Supeng Leng Jianhua He Ke Zhang School of Information and Communication Engineering University of Electronic Science and Technology of China(UESTC)Chengdu611731China Shenzhen Institute for Advanced Study UESTCShenzhen518000China School of Computer Science and Electronic Engineering University of EssexColchesterCO43SQUK

The Internet of Vehicles(Io V)has great potential for Intelligent Transportation Systems(ITS),enabling interactive vehicle applications,such as advanced driving and *** is crucial to ensure the reliability during the vehicle-to-vehicle interaction *** the emerging blockchain has superiority in handling security-related issues,existing blockchain-based schemes show weakness in highly dynamic Io *** the transaction broadcast and consensus process require multiple rounds of communication throughout the whole network,while the high relative speed between vehicles and dynamic topology resulting in the intermittent connections will degrade the efficiency of *** this paper,we propose a Digital Twin(DT)-enabled blockchain framework for dynamic Io V,which aims to reduce both the communication cost and the operational latency of *** address the dynamic context,we propose a DT construction strategy that jointly considers the DT migration and blockchain computing ***,a communication-efficient Local Perceptual Multi-Agent Deep Deterministic Policy Gradient(LPMA-DDPG)algorithm is designed to execute the DT construction strategy among edge servers in a decentralized *** simulation results show that the proposed framework can greatly reduce the communication cost,while achieving good security *** dynamic DT construction strategy shows superiority in operation latency compared with benchmark *** decentralized LPMA-DDPG algorithm is helpful for implementing the optimal DT construction strategy in practical ITS.

关键词： Digital twin Blockchain Internet of vehicles Multi-agent DDPG

来源：评论

学校读者我要写书评

暂无评论

Automatic Motorcyclist Helmet Rule Violation Detection using YOLOv5

Automatic Motorcyclist Helmet Rule Violation Detection using...

引用

2024 Asian Conference on Intelligent Technologies, ACOIT 2024

作者： Chheda, Meet Kapoor, Nupur Desai, Khushi Gupta, Rajeev Kumar School of Technology Computer Engineering Pandit Deendayal Energy University Gandhinagar India

ISBN: (纸本)9798350374933

Road traffic injuries rank among the leading causes of death worldwide, with motorcyclists accounting for a significant proportion of these fatalities. According to the World Health Organization, helmets are effective in reducing the risk of head injuries by approximately 69% and fatalities by about 42%. Despite these benefits, compliance with helmet laws remains inconsistent, underscoring the need for enhanced enforcement strategies. This paper presents the development of an Automatic Motorcyclist Helmet Rule Violation Detection system that employs the advanced YOLOv5 model to address this challenge. The system is designed to detect motorcyclists without helmets in real-time and simultaneously capture their vehicle number plates, thus aiding law enforcement in the effective enforcement of helmet laws. By automating the detection and reporting process, the system reduces the reliance on manual monitoring and helps ensure more consistent and unbiased enforcement across various road conditions and traffic volumes. Additionally, the real-time functionality of the system allows for immediate action to be taken against violators, potentially deterring future non-compliance and fostering a culture of safety among riders. Through this technology-driven approach, our project contributes to ongoing efforts in leveraging cutting-edge technology to improve road safety and reduce the incidence of motorcycle-related accidents and injuries, aligning with global initiatives for safer roads. © 2024 IEEE.

关键词： Highway accidents

来源：评论

学校读者我要写书评

暂无评论

Design and Implementation of Traffic Flow Prediction Model Based on Short and Long Time Memory Network 19th

Design and Implementation of Traffic Flow Prediction Model B...

引用

19th EAI International Conference on Heterogeneous Networking for Quality, Reliability, Security and Robustness, QShine 2023

作者： Liu, Sheng Li, Xinyue Cao, Ting Chang, Shuxiao School of Computer Science and Engineering Xi’an University of Technology Xi’an710048 China

ISBN: (纸本)9783031651250

Due to the randomness, fuzziness, time variability and uncertainty of traffic flow, it is difficult for traditional forecasting models based on time series or artificial neural networks to accurately reflect the actual traffic situation, etc. This paper takes the demand for short-term traffic flow forecasting of urban rail transit as the research object, and analyzes the implementation methods suitable for short-term traffic flow forecasting. LSTM neural network was used to construct the model for simulation experiment analysis. The results of data analysis show that the LSTM neural network model obtains the minimum average absolute percentage error MAPE value of 10.6% and the highest average accuracy of 89.4%, which has a good prediction effect and can improve the prediction work of short-term traffic flow. © ICST Institute for computer Sciences, Social Informatics and Telecommunications engineering 2024.

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

Analyzing Public Disputation of Autonomous Driving technology: A Study Based on Data Mining 4

Analyzing Public Disputation of Autonomous Driving Technolog...

引用

4th IEEE International Conference on Digital Twins and Parallel Intelligence, DTPI 2024

作者： Lv, Lehao Chen, Yuan Ma, Fanshu Li, Yongwei University of Science and Technology Beijing School of Computer and Communication Engineering Beijing China

ISBN: (纸本)9798350349252

With the rapid development of autonomous driving technology, public's opinion of this technology shows diversification and complexity. The discussion about autonomous driving online has a large quantity and diverse perspectives, which causes the traditional data analysis methods hard to catch and understand the whole spectrum of these data. This paper explores how to use text mining techniques effectively, especially the Latent Dirichlet Allocation (LDA) model, to analyze and understand people's attitude and viewpoint towards autonomous driving technology. By calculating the perplexity, coherence, Jensen-Shannon (JS)divergence and other related parameter indicators of the LDA model, striving to obtain the text mining model with best performance, in order to get valuable information faster and more efficiently. © 2024 IEEE.

关键词： Autonomous driving data mining Latent Dirichlet Allocation

来源：评论

学校读者我要写书评

暂无评论

Video Based Cricket Shot Classification using modified LRCN Approach

Video Based Cricket Shot Classification using modified LRCN ...

引用

2024 IEEE Silchar Subsection Annual Conference, SILCON 2024

作者： Krithiga, R. Ramesh, Balaji Sriraam, C. Lionel Donato, L. Rojesh, S. School of Computer Science and Engineering Vellore Institute of Technology TamilNadu Chennai India

ISBN: (纸本)9798331540821

In recent years, advancements in Artificial Intelligence (AI) have caused major transformations in the realm of cricket analysis. From player performance evaluation to strategic decision-making, AI technologies have revolutionized how cricket is understood and played. In this paper, we focus on the task of classifying cricket shots using video data, leveraging deep learning methodologies. The proposed model aims to classify cricket videos into five distinct shot categories, utilizing the CricShot10 dataset originally comprising ten shot classes. With approximately 100-150 video clips per shot type, we streamline the classification task by selecting five fundamental shots for analysis. To accomplish this, we propose a modified architecture of the popular architecture combining Convolutional Neural Networks (CNNs) with Long Short-Term Memory (LSTM) networks, commonly referred to as Long-Term Recurrent Convolutional Network (LRCN). Additionally, we integrate techniques such as MaxPooling2D, Dropout, Flatten, and Dense layers to enhance the model's performance and robustness. Through this proposal, we not only contribute to the field of cricket analysis but also demonstrate the efficacy of deep learning approaches in video-based sports analytics. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

A Client Detection and Parameter Correction Algorithm for Clustering Defense in Clustered Federated Learning 30

A Client Detection and Parameter Correction Algorithm for Cl...

引用

30th International Conference on Mobile Computing and Networking, ACM MobiCom 2024

作者： Ye, Junyu Shi, Lei Xu, Hao Pan, Sinan Xu, Juan School of Computer Science and Information Engineering Hefei University of Technology HeFei China

ISBN: (纸本)9798400704895

As a new federated learning(FL) paradigm, clustered federated learning (CFL) could effectively address the issue of model training accuracy loss due to different data distribution in FL. However, the introduction of the clustering process also brings new risks. Adversaries can implement model poisoning by adding crafted perturbations with clients' model parameters, potentially resulting in overall clustering failure. To tackle this problem, we propose a client detection and parameter correction framework in this paper. Our approach aims to identify malicious clients by analyzing the difference in vector parameter density distribution between malicious and benign clients. We precisely locate malicious perturbations in the parameters and recover them, enabling the server to effectively utilize benign updates for normal clustering and training within the CFL framework. Experiment results show that our defense algorithm outperforms others, consistently improving training accuracy by an average of 30% under various kinds of attacks. © 2024 Copyright is held by the owner/author(s). Publication rights licensed to ACM.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Convolution Neural Network Regression Model to Predict Personality Scores

引用

Journal of engineering Science and technology Review 2022年第4期15卷 25-32页

作者： Bhamare, Mamta Ashokkumar, K. Department of Computer Science and Engineering Sathyabama Institute of Science & Technology Chennai India School of Computer Engineering and Technology MITWPU Pune India

The world has witnessed tremendous increase in the use of social media platforms in the last decade. Posting and updating on such platforms have become a way for people to express their opinions, experiences, feelings, views etc. Along with all these, they tend to post and share their personal Information. This Information can be used in advantageous way to understand their needs and requirements The approach, known as personality prediction, entails sepΑarating digital input into attributes and mapping it according to a personality model. In this paper, we are interested to predict the personality of such users. Researchers have shown based on the social behavior towards friends, number of friends, groups joined etc. can help in predicting the personality trait from Facebook and Twitter. The text posts for the user are converted to be suitable representation using GloVe algorithm. And then feeding it to Convolution Neural Net, a popular Deep Learning architecture on the text posts that was implemented for the traits present in the Big Five trait theory. This method helped to achieve comparable outputs with other recent models © 2022 school of Science, IHU. All rights reserved

关键词： Convolution

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：