检索结果-内蒙古大学图书馆

41st computer Graphics international conference, CGI 2024

作者： Wu, Tong Wang, Song Peng, Xin School of Computer Science and Technology Southwest University of Science and Technology Mianyang China

ISBN: (纸本)9783031820205

The paper presents a recommender algorithm for visual analysis based on Data field Schema and Aggregation, and developed an automated data analysis solution recommendation system (AutoEDA) in conjunction with the Exploratory Data Analysis process (EDA). To address the issue of field type recognition, we designed a dynamic field recognition framework combining four classification models to automatically classify data field types. For the evaluation of analysis strategies, we proposed an analysis strategy evaluation method based on the Chart Attribute Length for designing custom search strategies and defining chart matching rules. Finally, through interactive visual backtracking and the functionality of modifying and tracing analysis solutions, we collaborated with data analysts of varying experience levels, and validated the effectiveness and feasibility of the proposed solution with structured data from publicly available websites. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Data aggregation

来源：评论

学校读者我要写书评

暂无评论

Multi-Scene Fire Smoke Detection Algorithm Based on YOLO v8 24

Multi-Scene Fire Smoke Detection Algorithm Based on YOLO v8

引用

14th international conference on Communication and network Security, ICCNS 2024

作者： Li, Aiguo Song, Xiaowan School of Computer Science and Technology Xi'an University of Science and Technology Shaanxi Xi'an China

ISBN: (纸本)9798400717994

Aiming at the problems of semi-transparent smoke misdetection and fire missing detection in multi-scene fire smoke detection, a multi-scene fire smoke detection algorithm based on YOLOv8 is proposed. Firstly, deformable convolution DCN V2 is introduced to solve the problems of great difference in size and shape of flame target and high difficulty in detection, so as to achieve full feature extraction. Secondly, dynamic sparse attention is used to filter out irrelevant key-value pairs at the rough area level, so as to enhance the focus on fire. Finally, by fusing generalized feature pyramid GFPN to enrich the multi-scale feature sensing field, further improve the performance of the algorithm and save computational resources. Experimental results show that the precision rate of YOLOv8-FS algorithm is 94.7% mAP50 reaches 96.1% and the model size is only 7.2 MB. The algorithm can balance precision and lightweight, and can meet the requirements of fire detection in complex scenes. © 2024 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Smoke detectors

来源：评论

学校读者我要写书评

暂无评论

Cyber-Physical Systems and IoT Intrusion Detection: A Hybrid Deep Learning Perspective 3

Cyber-Physical Systems and IoT Intrusion Detection: A Hybrid...

引用

3rd international conference on Electronics and Renewable Systems, ICEARS 2025

作者： Prakash, S. Jaya Devi, D. Muthukumaran, N. Romila, R. Idhaya Engineering College for Women Department of Computer Science and Engineering Tamilnadu Chinnasalem India Maha Barathi Engineering College Department of Computer Science and Engineering Tamilnadu Chinnasalem India A.R. Engineering College Department of Computer Science and Engineering Tamilnadu Villupuram India Sri Muthukumaran Institute of Technology Department of Information Technology Tamilnadu Chennai India

ISBN: (纸本)9798331509675

Several vital resources are increasingly being protected by cyber-physical systems (CPSs), makes the detection of incidents on these systems critical. CPSs along with other domains, such as the Internet of Things (IoT), frequently use machine learning (ML) and deep learning (DL) approaches to combat intrusion detection. Existing approaches, however, frequently favor the deployment of advanced detection models over their usefulness in actual operations. This research work develops a hybrid intrusion detection technique by combining Convolutional Neural networks (CNN) with Long Short-Term Memory (LSTM). This work analyzes the experimental results to LSTM and CNN to develop an effective comparative approach. The experiments are conducted on the two datasets. The efficiency indicators examined are intrusion detection results such as accuracy, precision, recall, and f1score. This comparison indicates that the proposed hybrid algorithm is the preferable approach. © 2025 IEEE.

关键词： network intrusion

来源：评论

学校读者我要写书评

暂无评论

Adaptive Crow Search Algorithm for Hierarchical Clustering in Internet of Things-Enabled Wireless Sensor networks

引用

international Journal of Advanced computer science and Applications 2025年第4期16卷 533-541页

作者： Wang, Lingwei Wang, Hua Guangdong University of Science and Technology Guangdong Dongguan523083 China

The Internet of Things (IoT) relies on efficient Wireless Sensor networks (WSNs) for data collection and transmission in various applications, including smart cities, industrial automation, and environmental monitoring. Clustering is a fundamental technique for structuring WSNs hierarchically, enabling load balancing, reducing energy consumption, and extending network lifespan. However, clustering optimization in WSNs is an NP-hard problem, necessitating heuristic and metaheuristic approaches. This study introduces an Adaptive Crow Search Algorithm (A-CSA) for clustering in IoT-enabled WSNs, addressing the inherent limitations of the standard CSA, such as premature convergence and local optima entrapment. The proposed A-CSA incorporates three key enhancements: 1) a dynamic awareness probability to improve global search efficiency during initial population selection, 2) a systematic leader selection mechanism to enhance exploitation and avoid random selection bias, and 3) an adaptive local search strategy to refine cluster formation. Performance evaluations conducted under varying network configurations, including node density, network size, and base station positioning, demonstrate that A-CSA outperforms existing clustering approaches in terms of energy efficiency, network longevity, and data transmission reliability. The results highlight the potential of A-CSA as a robust optimization technique for clustering in IoT-driven WSN environments. © (2025), (science and Information Organization). All Rights Reserved.

关键词： Sensor nodes

来源：评论

学校读者我要写书评

暂无评论

A Comparative study on facial expression recognition using MobileNetV2, VGG-16, ResNet and swin transformer 3

A Comparative study on facial expression recognition using M...

引用

3rd international conference on Algorithms, network, and Communication technology, ICANCT 2024

作者： Bi, Kairui Liao, Jiapei Gao, Dongting Yin, Ruilin Wang, Yize Semiahmoo Secondary School BC Canada Faculty of Science and Technology University of Macau 999078 China College of Science and Technology Temple University PhiladelphiaPA19122 United States Division of Science and Technology Beijing Normal University Hong Kong Baptist University United International College Zhuhai519087 China College of Engineering and Physical Sciences University of Birmingham Edgbaston BirminghamB15 2TT United Kingdom

ISBN: (数字)9781510688902

ISBN: (纸本)9781510688896

In the modern era, the implementation of computer vision technologies has become increasingly prevalent across various domains. In many fields, such as security and identification, computer vision technologies are widely implemented and have gradually become increasingly prevalent topics. Recent advancements in facial emotion recognition models have proved to be significant for the development of different deep learning models in past research. Undoubtedly, Convolutional Neural networks (CNNs) have long been the dominant approach for such facial recognition tasks, especially for emotion classification tasks. Instead of the traditional CNNs, some other researchers prefer using vit such as (Mobile Swim Windows Attention) to solve such problems because they argue that it could extract more global features rather than local features extracted by CNN. To determine the advantages and disadvantages on these models for the performance tasks, this paper presents a comparative analysis of CNN-based models - MobileNetV2, VGG16, ResNet - against the Swin Transformer, a Vision Transformer (ViT) model. Using the FER2013 datasets with 48x48 pixels images that are all classified into 7 different emotion categories and implementing the augmentation techniques to improve the quality of the datasets, we utilized the pretrained models, train them with at least 20 epochs, and evaluated the performance of classical convolutional neural networks and Vision Transformer on this dataset based on accuracy, F1 Score, Precision, and Recall with accuracy, loss, and confusion matrix visualization. © 2025 SPIE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Image Segmentation network Based on Convolutional Layers, a Clustering Layer and Cluster Classification 37th

Image Segmentation Network Based on Convolutional Layers, a ...

引用

37th international conference on computer Applications in Industry and Engineering

作者： Rosas-Romero, Roberto Univ Amer Puebla San Andres Cholula Puebl Mexico

ISBN: (纸本)9783031762727;9783031762734

This work contributes to the detection of regions of interest on images and their corresponding classification in medical imaging applications by introducing an image segmentation network that consists of four stages. In the first stage, multi-resolution processing is applied to outline regions where further segmentation and classification are to be conducted. Subsequently, a quad-tree division stage followed by a clustering stage deliver an image that is divided into unlabeled clusters. The output stage assigns each cluster to one class. This architecture offers flexibility in the input and output stages since this network can be fed with images of any size and the output stage can be implemented with any traditional classification model such as k-nearest neighbors, multi-layer perceptron, and support vector machine. Another contribution of this work is that this network does not rely on a very large number of annotated images for its training.

关键词： Image segmentation Convolutional Neural networks Multi-Resolution Analysis Clustering

来源：评论

学校读者我要写书评

暂无评论

LowResIR-Net: Lightweight Residual network for ADL Recognition Using Low-Resolution Infrared Sensor Data 16

LowResIR-Net: Lightweight Residual Network for ADL Recogniti...

引用

16th international conference on ENTERprise Information Systems, CENTERIS 2024 - 12th international conference on Project MANagement, ProjMAN 2024 - 14th international conference on Health and Social Care Information Systems and Technologies, HCist 2024

作者： Mekruksavanich, Sakorn Phaphan, Wikanda Jitpattanakul, Anuchit Department of Computer Engineering School of Information and Communication Technology University of Phayao Phayao56000 Thailand Research Group in Statistical Learning and Inference Department of Applied Statistics Faculty of Applied Science King Mongkut's University of Technology North Bangkok Bangkok10800 Thailand Intelligent and Nonlinear Dynamic Innovations Research Center Department of Mathematics Faculty of Applied Science King Mongkut's University of Technology North Bangkok Bangkok10800 Thailand

Detecting activities of daily living (ADL) is crucial for supported living and medical monitoring. Traditional approaches rely on high-resolution sensors and computationally intensive algorithms, limiting their scalability and practical use in real-world settings. This research introduces LowResIR-Net, a compact deep residual network architecture specifically tailored for recognizing ADL utilizing low-resolution infrared sensor data. LowResIR-Net efficiently extracts key features from low-dimensional data by leveraging the advantages of infrared sensing, such as privacy protection and light resilience. Its architecture combines residual connections with depth-wise convolutions, enhancing feature learning while maintaining a compact model size. The model’s performance is evaluated using the Coventry-2018 dataset, containing low-resolution infrared data from three residential sensor locations. Our tests have shown that in all three cases, LowResIR-Net consistently achieves superior performance compared to other cutting-edge deep learning models, such as CNN, LSTM, BiLSTM, GRU, and BiGRU. LowResIR-Net attains the most excellent accuracies of 90.18%, 93.75%, and 93.30%, and F1-scores of 81.81%, 89.55%, and 88.92% for scenarios I, II, and III, respectively. Furthermore, it achieves these results with fewer parameters than most other models. LowResIR-Net’s strong performance highlights its ability to accurately capture complex patterns in low-resolution infrared data for ADL detection. This study demonstrates the potential of lightweight deep learning and low-resolution infrared sensors in developing privacy-focused and computationally efficient wellness monitoring systems. © 2025 The Author(s).

关键词： Infrared detectors

来源：评论

学校读者我要写书评

暂无评论

Application research on construction safety management of building engineering based on computer vision technology 24

Application research on construction safety management of bu...

引用

5th international conference on computer science and Management technology, ICCSMT 2024

作者： Han, Yuanyuan School of Humanities Shanghai Institute of Technology Shanghai China

ISBN: (纸本)9798400709999

With the rise of artificial intelligence, the application of computer vision technology in the construction safety management of construction engineering is increasingly *** the application of computer vision technology in the construction safety management is not only an upgrade of the existing safety management mode, but also provides theoretical contributions and technical support for the realization of future intelligent ***, this study discusses the application of computer vision technology in the construction safety management of construction projects by constructing a construction engineering safety control *** framework evaluates the construction safety situation reasonably and quantitatively through the processing process of image acquisition and preprocessing, image recognition, data analysis and control system on the construction *** the visual management of construction site personnel, construction equipment, construction materials, construction standards and construction environment, and improve the safety management effect of construction site. © 2024 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Project management

来源：评论

学校读者我要写书评

暂无评论

Enhancing Motor Imagery Classification with Residual Graph Convolutional networks and Multi-Feature Fusion

引用

international JOURNAL OF NEURAL SYSTEMS 2025年第1期35卷 2450069页

作者： Xu, Fangzhou Shi, Weiyou Lv, Chengyan Sun, Yuan Guo, Shuai Feng, Chao Zhang, Yang Jung, Tzyy-Ping Leng, Jiancai Qilu Univ Technol Shandong Acad Sci Int Sch Optoelect Engn Jinan 250353 Peoples R China Shandong Univ Qilu Hosp Rehabil Ctr Jinan 250012 Peoples R China Shandong Univ Tradit Chinese Med Affiliated Hosp Rehabil & Phys Therapy Dept Jinan 250011 Peoples R China Univ Calif San Diego Inst Neural Computat Swartz Ctr Computat Neurosci La Jolla CA USA

Stroke, an abrupt cerebrovascular ailment resulting in brain tissue damage, has prompted the adoption of motor imagery (MI)-based brain-computer interface (BCI) systems in stroke rehabilitation. However, analyzing electroencephalogram (EEG) signals from stroke patients poses challenges. To address the issues of low accuracy and efficiency in EEG classification, particularly involving MI, the study proposes a residual graph convolutional network (M-ResGCN) framework based on the modified S-transform (MST), and introduces the self-attention mechanism into residual graph convolutional network (ResGCN). This study uses MST to extract EEG time-frequency domain features, derives spatial EEG features by calculating the absolute Pearson correlation coefficient (aPcc) between channels, and devises a method to construct the adjacency matrix of the brain network using aPcc to measure the strength of the connection between channels. Experimental results involving 16 stroke patients and 16 healthy subjects demonstrate significant improvements in classification quality and robustness across tests and subjects. The highest classification accuracy reached 94.91% and a Kappa coefficient of 0.8918. The average accuracy and F1 scores from 10 times 10-fold cross-validation are 94.38% and 94.36%, respectively. By validating the feasibility and applicability of brain networks constructed using the aPcc in EEG signal analysis and feature encoding, it was established that the aPcc effectively reflects overall brain activity. The proposed method presents a novel approach to exploring channel relationships in MI-EEG and improving classification performance. It holds promise for real-time applications in MI-based BCI systems.

关键词： Brain network modified residual graph convolutional network modified S-transform self-attention mechanism stroke

来源：评论

学校读者我要写书评

暂无评论

network Traffic Prediction using Deep Learning Models: Reducing Latency in High-Demand Scenarios

Network Traffic Prediction using Deep Learning Models: Reduc...

引用

2025 international conference on Computing for Sustainability and Intelligent Future, COMP-SIF 2025

作者： Sagar, B.J. Job Karuna Divya, M.V. Kale, Namita Anand, Neetu Gaurav, Kumar Shashidhara, Shrividya Joginpally B.R Engineering College Department of Cse Telangana Hyderabad India East West Institute of Technology Department of Computer Science Bangalore India MET'S Institute Of Engineering Department of Information Technology Nashik India Maharaja Surajmal Institute Department of Computer Applications Delhi India Bms Institute of Technology and Management Department of Cse Bangalore India

ISBN: (纸本)9798331538538

The availability of the internet and enhanced applications, more specifically, the utilization of more devices in networks, and improved computational applications has created complexity in the control of network traffic. network traffic forecasting is an essential component in low-latency networks, as well as in real-time applications that deserve excellent service quality and efficient resource allocation. This study therefore focuses on identifying whether deep learning models can further improve the accuracy in forecasting traffic flows in a network. Therefore, with the help of LSTM networks and Convolutional Neural networks, the work will build temporal dependencies and spatial correlations of network traffic data. There are major and minor adaptations to some classic methodologies that conform the proposed framework;moreover, it is applied to two real datasets to highlight the capacity of the model in anticipating traffic congestion with a higher level of accuracy than the traditional methodologies. Evaluation of models in dynamic network conditions is done through common parameters such as prediction accuracy, lowered latency, and scalability. Saved models show that deep learning-based solutions reduce latency at high loads much more effectively, which will help network administrators in finding efficient solutions. This study demonstrate how deep learning can aid in the prediction of traffics, and the manner in which it is changing the efficiency and effectiveness of modern-day network systems. It makes the foundation for improved understanding of traffic patterns in order to apply it to optimize future generation of communication networks. © 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

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：