检索结果-内蒙古大学图书馆

Informatica (Slovenia) 2024年第16期48卷 27-36页

作者： Gao, Li Qiu, Junlin Chen, Guanhua Faculty of Computer and Software Engineering Huaiyin Institute of Technology Huai’an223003 China

As software development models and methods mature, large-scale software systems emerge. However, a critical challenge remains: the lack of a comprehensive software test data management model that integrates basic data management with advanced knowledge reasoning. To address this issue, we developed a software test data management model based on knowledge graphs, enabling intelligent management and reasoning of software test data. The model incorporates an entity extraction model based on a feed-forward neural network, a knowledge graph integration method based on graph databases, and a knowledge reasoning submodule based on deep learning. To validate the effectiveness of our model, we evaluated the performance of each component individually. Our deep learning-based entity extraction model achieved an accuracy of 0.92, a recall of 0.88, and an F1 score of 0.90, significantly outperforming traditional methods such as regular expressions and dictionary-based approaches. Utilizing Cypher for graph database querying, our system provides accurate answers with a response time of 0.12 seconds, outperforming SQL and SPARQL-based querying methods. Furthermore, our approach excels in knowledge-based reasoning with an accuracy of 0.89 and site coverage of 0.81, surpassing both ontology-based and graph-based reasoning methods. These results highlight the enhanced construction, querying, and reasoning capabilities of our knowledge graph-based approach for managing software testing data. © 2024 Slovene Society Informatika. All rights reserved.

关键词： Knowledge graph

来源：评论

学校读者我要写书评

暂无评论

Enhancing the Evaluation Performance of Convolutional Neural Networks-Based Vehicle Classification Systems

Enhancing the Evaluation Performance of Convolutional Neural...

引用

2023 International Conference on Cognitive Computing and Complex Data, ICCD 2023

作者： Kiyindou, Michel Precieux Sunday, Samuel Enobong Hong, Zhou Huaiyin Institute of Technology Faculty of Electronic Information Engineering Huai'an City China Huaiyin Institute of Technology Foreign Expeert Workshop Huai'an Faculty of Computer Engineering Huai'an City China

ISBN: (纸本)9798350306293

The rapid advancement of artificial intelligence, particularly deep learning, has spurred its integration into diverse domains. Among these, the application of deep learning in vehicle classification stands out as a vital focus within intelligent transportation systems, road traffic planning, safety warning, and driverless driving. In this research, a comprehensive approach of CNN's evolution such as LeNet, AlexNet, and other CNN-based vehicle classification techniques are enhanced for a better performance, while also investigating R-CNN, YOLO, and related methods for vehicle detection. Evaluative datasets for these methods are highlighted. Moreover, the enhancement of the computing power of the Graphics Processing Unit (GPU) and the sharp increase in the amount of data, deep learning based on convolutional neural networks has become a research hotspot. To evaluate the accuracy and efficiency of our vehicle detection methods, RCNN and YOLO series datasets were used to evaluate these methods and an the development of CNN as the main line was introduced and use to analyze the vehicle classification methods based on LeNet, AlexNet and other CNNs. This research is pivotal for vehicle tracking, traffic management, and autonomous driving progress. In this landscape, convolutional neural networks (CNNs), self-encoder neural networks, and deep belief networks take prominence, with CNNs evolving as the most dynamic. Notably, CNN's development has led to refined vehicle classification through models like LeNet, AlexNet, VGG, and GoogLeNet. The paper concludes by outlining prospective research directions in this domain. © 2023 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Human Pose Estimation for Boxing Based on Model Transfer Learning 4

Human Pose Estimation for Boxing Based on Model Transfer Lea...

引用

4th International Conference on High Performance Big Data and Intelligent Systems, HDIS 2022

作者： Lin, Jianchu Xie, Xiaolong Wu, Wangping Xu, Shengpeng Liu, Chunyan Hudoyberdi, Toshboev Huaiyin Institue of Technology Faculty of Computer and Software Engineering Huai'an China

ISBN: (数字)9781665491440

ISBN: (纸本)9781665491440

The boxing as a competitive sport is now walking into Chinese campus and daily fitness exercises. The rapid growth of boxing sports has resulted in a coaching shortage. The pose of boxing can be estimated by using human pose estimation technology in artificial intelligence, to teach interns and relieve the coach shortage. However, the large number of training datasets and the lack of consistency of camera views in training limit the technology's practical application. Therefore, the model transfer methodology was implemented to address these limitations, three basic models of OpenPose (OP), High Resolution (HR), and stacked Hourglass (HG) networks were employed for model transfer learning and analyzing of boxing application. The results show the model transfer can efficiently improve the average accuracy of pose key points from 1% to 19%, model transfer can practically help address the issue of dependence on dataset collection. The HR network achieves the best performance among these three networks, but the HG network still gets better estimation on some key points. © 2022 IEEE.

关键词： Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

A deep neural network based feature fusion panoramic segmentation method

A deep neural network based feature fusion panoramic segment...

引用

International Conference on Emerging Electronic and Automation technology 2022, ICEEAT 2022

作者： Fei, Hongyan Mokhtarzadeh, Amir Ali Zhou, Jiayu Geng, Junzuo Huaiyin Institute of Technology School of Computer and Software Engineering 223003 China Jiangsu Foreign Expert WorhShop Jiangsu Huaian223003 China Nanjing Institute of Technology School of Computer and Software Engineering 211167 China

Segmentation of complex and diverse scenes in the field of machine vision has the advantages of low cost and good robustness. However, in the real open intelligent driving scene, the segmented object will be affected by the change of proportion and illumination. The feature extraction of the interference algorithm causes some problems such as local information loss and feature background fusion in the process of execution. In this paper, deep neural network method is used for feature fusion panoramic segmentation. It focuses on the premise of lightweight, and improves the performance by improving the structure of semantic segmentation network layer and model calculation effect, and reducing the number of network parameters and calculations, so as to further improve the detection performance. At the same time, with the gradual deepening of deep learning network architecture, but most embedded devices can not bear a huge amount of computing workload. An attentional channel pruning based approach is proposed using a wide range of experimental and ablation studies to compare different data, including VOC2012 and ImageNet datasets. We show that the proposed method outperforms state of the art methods in terms of visual quality. © Published under licence by IOP Publishing Ltd.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

A Method for Depth Camera Calibration Based on Motion Capture System

A Method for Depth Camera Calibration Based on Motion Captur...

引用

2023 International Conference on Cognitive Computing and Complex Data, ICCD 2023

作者： Lin, Jianchu Zhao, Ming Yin, Guojun Zhou, Haiping Hudoyberdi, Toshboev Jiang, Bo Huaiyin Institue of Technology Faculty of Computer and Software Engineering Huai'an China Huaiyin Institue of Technology Faculty of Art and Design Huai'an China Jiangsu Eazytec Co. Ltd Yixin China

ISBN: (纸本)9798350306293

With the rapid development of 3D computer vision technology, depth cameras have been widely used. Depth camera calibration is mainly constrained by two aspects: (1) alignment of depth data, and (2) distortion in RGB images. Under the condition of unknown camera intrinsic parameters, a method for RGBD commercial camera calibration is studied based on an optical motion capture system. The proposed approach utilizes the projection from the motion capture space to the depth camera's field of view, establishing a coupled parameter model for the camera's CCD photosensitive elements to improve the calibration of the depth camera. Experimental results demonstrate a 33% reduction in errors in the central region of the camera's field of view compared to commonly Least Squares (LS) calibration methods. Moreover, the method exhibits faster computation speed, enabling efficient real-time calibration of depth cameras with unknown parameters in practice. © 2023 IEEE.

关键词： CCD cameras

来源：评论

学校读者我要写书评

暂无评论

Optimization Model for Garbage Image Classification Based on RepVGG

Optimization Model for Garbage Image Classification Based on...

引用

Cognitive Computing and Complex Data (ICCD), International Conference on the

作者： Ken Lin Shaozhang Xiao Kaikai Shang Faculty of Computer and Software Engineering Huaiyin Institute of Technology Huaian China

ISBN: (数字)9798350352894

ISBN: (纸本)9798350352900

The application of deep learning to the classification of garbage images has been demonstrated to significantly enhance the accuracy and efficiency of classification through the extraction of features and the recognition of patterns. However, in environments with limited resources or computational power, traditional models are unable to effectively handle complex trash image classification tasks. Consequently, the deployment of lightweight models in garbage image classification h as g arnered increasing interest. In this paper, the structure of the RepVGG model is optimized by combining depthwise separable convolution. Adapting the convolutional layers of the original trunk structure to training branches and replacing the convolutional kernel of the trunk with a 5 × 5 convolutional kernel results in the trunk and branches being reparameterized to a single 5 × 5 convolutional layer after training. This design enhances the image feature extraction capability at multiple levels. The inference structure retains the simple stacking of the original convolutional and ReLU layers. In this paper, the improved model is named DSC5RepVGG. The experimental results demonstrate that DSC5RepVGG reduces the number of parameters by 60.97% in comparison to RepVGG, reduces the theoretical floating-point computation by 58.22%, and improves the Top-1 accuracy by 1.29% on Kaggle’s Garbage Classification dataset.

关键词： Training Adaptation models Accuracy Convolution Computational modeling Stacking Feature extraction Kernel Image classification Periodic structures

来源：评论

学校读者我要写书评

暂无评论

Efficient Privacy-Preserving Federated Learning with Byzantine Robustness

Efficient Privacy-Preserving Federated Learning with Byzanti...

引用

Cognitive Computing and Complex Data (ICCD), International Conference on the

作者： Chunhua Jin Lulu Li Jiahao Wang Youwang Sun Ying Jin Guanhua Chen Faculty of Computer and Software Engineering Huaiyin Institute of Technology Huai’an China

ISBN: (数字)9798350352894

ISBN: (纸本)9798350352900

Federated learning is a distributed machine learning paradigm. It entails training a global model across diverse organizations while ensuring privacy constraints are upheld. Despite its promising potential, challenges arise when adversaries attempt to infer private information from exchanged parameters or compromise the global model. Although the development of multiple protocols to mitigate security threats, safeguarding the privacy of individual participants while countering Byzantine adversaries remains a challenge. Furthermore, the frequent transmission of gradients between the master server and participants results in significant communication overhead, thereby limiting the training efficiency of federated learning. In this study, We propose an efficient privacy-preserving federated learning scheme with Byzantine robustness (EFL-PB). For privacy protection and Byzantine robustness, we propose a novel aggregation strategy based on the multi-message shuffle protocol in differential privacy. For enhancing communication efficiency, we develop an adaptive gradient compression scheme. Theoretical analysis demonstrates that our multi-message shuffle protocol can satisfy differential privacy. Experimental results on the MNIST, FashionMNIST, and CIFAR-10 datasets validate the robustness and efficiency of the EFL-PB scheme. As well, the EFL-PB scheme exhibits strong performance on non-iid datasets.

关键词： Training Privacy Differential privacy Protocols Federated learning Resists Robustness Security Servers Protection

来源：评论

学校读者我要写书评

暂无评论

Garbage Image Classification Based on Improved ShuffleNet Fusion Algorithm

Garbage Image Classification Based on Improved ShuffleNet Fu...

引用

Cognitive Computing and Complex Data (ICCD), International Conference on the

作者： Kaikai Shang Shaozhang Xiao Ken Lin Kangliang Xiao Yin Xia Faculty of Computer and Software Engineering Huaiyin Institute of Technology Huaian China

ISBN: (数字)9798350352894

ISBN: (纸本)9798350352900

Aims to address the problems of low accuracy of existing garbage classification and low classification ac curacy of models, large model size and difficulty of deployment on portable devices. A lightweight garbage image classification model is proposed by merging the improved lightweight model ShuffleNet V2 and MobileViT. Firstly, the structure of the ShuffleNet V2 model is adjusted and improved to enhance the cross-channel information interaction ability of the model, which improves the feature information interaction ability between the channels of the ShuffleNet V 2 model; secondly, the improved ShuffleNet V2 model is fused with the MobileViT model, the advantage of the MobileViT module in global feature information extraction is used to overcome the shortcomings of the ShuffleNet V2 network, which is difficult to extract global feature information. Comparative experiments are conducted on public datasets to evaluate the performance of the model. The experimental results show that the accuracy of the proposed SNViT model is significantly better than the D enseNet-121 model, and also has a better classification performance compared to EfficientNet V2, MobileNet V2 and other models, while the precision rate, recall rate and F1 value are also improved. The improved SNViT model improves the accuracy of the image classification task and achieves high classification performance, while the number of parameters is small and easy to train and deploy.

关键词： Accuracy Convolution Computational modeling Merging Feature extraction Information retrieval Data models Data mining Image classification Periodic structures

来源：评论

学校读者我要写书评

暂无评论

Robot positioning and navigation technology is based on Integration of the Global Navigation Satellite System and real-time kinematics

Robot positioning and navigation technology is based on Inte...

引用

International Conference on Emerging Electronic and Automation technology 2022, ICEEAT 2022

作者： Geng, Junzuo Mokhtarzadeh, Amir Ali Fei, Hongyan Gao, Han Ding, Nan Kong, Rongshuang Zhou, Jiayu Huaiyin Institute of Technology School of Computer and Software Engineering 223003 China Jiangsu Foreign Expert WorhShop Jiangsu Huaian223003 China Nanjing Institute of Technology School of Computer and Software Engineering 211167 China

This paper provides a new robot positioning and navigation technique that combines GNSS-RTK and visual SLAM to address the shortcomings of existing single robot positioning and navigation techniques, such as limited adaptation to scenarios and susceptibility to environmental influences. Real-time kinematics (RTK) is a real-time dynamic observation technique based on carrier phase observations, which can provide real-time 3D positioning results of the measurement site in a specified coordinate system, and can be positioned with centimetre-level accuracy, which is suitable for It is suitable for outdoor positioning of mobile robots. In this paper, a dataset is studied to compare and analyse the advantages and disadvantages of different SLAMs in order to determine the most suitable robot positioning and navigation technique for autonomous indoor and outdoor mobility. These sensors were well calibrated and the composition was applied to the RK3588. The results show good performance in indoor-outdoor switching scenarios, obtaining a mapping technique that cannot be achieved with a single visual SLAM, effectively improving the ability to adapt to multiple scenarios and enhancing the operational range of the transportable robot. © Published under licence by IOP Publishing Ltd.

关键词： Kinematics

来源：评论

学校读者我要写书评

暂无评论

Intent Classification of Nepali Factoid Questions

Intent Classification of Nepali Factoid Questions

引用

2023 International Conference on Cognitive Computing and Complex Data, ICCD 2023

作者： Timilsina, Sushil Khadka, Ashim Lin, Jianchu Faculty of Science and Technology Gandaki College of Engineering and Science Pokhara Nepal Faculty of Science and Technology Nepal College of Information Technology Lalitpur Nepal Faculty of Computer and Software Engineer Huaiyin Institute of Technology Jiangsu China

ISBN: (纸本)9798350306293

Question Classification plays a vital role in identifying the accurate answer to any question, and is considered as a core component of Question Answering Systems. In Nepali Natural Language Processing, the area of Question Answering has not even been initiated due to a lack of dataset. The dataset of Nepali factual questions has been created by collecting questions from general knowledge books and labelling them with developed taxonomy. The problem of Nepali factoid question classification has been studied with a Support Vector Machine, which works as a baseline classification algorithm and obtained a 0.74 weighted F1-score in the imbalance dataset. The experimental results of the proposed model show the effectiveness of the approach in classifying factoid questions accurately compared with the state-of-the-art approaches. © 2023 IEEE.

关键词： Natural language processing systems

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：