检索结果-内蒙古大学图书馆

Enhanced deep learning framework for real-time pain assessment using multi-modal fusion of facial features and video streams

引用

engineering Applications of Artificial Intelligence 2025年 154卷

作者： El-Ghaish, Hany Al-Basiouny, Mohamed Yousry Alshewimy, Mahmoud A.M. Tanta University Faculty of Engineering Computer and Control Engineering Dept. Tanta Egypt

This paper introduces an enhanced deep learning multimodal framework for real-time pain assessment that addresses the limitations of the single-mode approaches. Specifically, time-related disadvantages and its poor temporal analysis, noted in recent studies. The proposed framework fuses VGG16 (Visual Geometry Group 16-layer network), a pre-trained convolutional neural network (CNN), with a bidirectional long short-term memory network (Bi-LSTM) augmented with attention mechanisms. VGG16 has a proven ability to capture fine-grained spatial information in facial expressions to extract powerful spatial features, whereas the Bi-LSTM augmented with attention mechanisms because of its computational efficiency and concisely captures temporal features. Moreover, adaptive boosting has used seeking generalization across a broad base of settings and Gradient-weighted Class Activation Mapping (Grad-CAM) to provide comprehensible and interpretable visual explanations of the decision-making processes of the framework. Extensive evaluations were conducted on diverse popular benchmark datasets: BioVid, UNBC-McMaster, and MIntPAIN. Our framework achieves an AUC (the Area under the Receiver Operating Characteristic Curve) of 0.99 and 99% cross-validation accuracy, corresponding to a 12.5% improvement in accuracy and a 20% reduction in false positives relative to state-of-the-art convolutional neural network–long short-term memory (CNN–LSTM) models. Real-time deployment on an NVIDIA Jetson Nano yields 120 ms (ms) inference per frame with less than 1% accuracy loss, demonstrating its feasibility for edge-based telemedicine. Combined with explainability via Grad-CAM and privacy-preserving vision-only inputs, this framework offers an ethical, scalable solution for nonverbal pain detection. © 2025 Elsevier Ltd

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Semantic Communication based on Generative AI: A New Approach to Image Compression and Edge Optimization

arXiv

引用

arXiv 2025年

作者： Pezone, Francesco Barbarossa, Sergio Caire, Giuseppe Di Lorenzo, Paolo Kountouris, Marios Bennis, Mehdi Sapienza University of Rome Department of Computer Control and Management Engineering Italy Technische Universität Berlin Faculty 4 - Electrical Engineering and Computer Science Germany

As digital technologies continue to advance, modern communication networks face unprecedented challenges in handling the vast amounts of data produced daily by connected intelligent devices. Autonomous vehicles, smart sensors, IoT systems etc., are gaining more and more interest and new communication paradigms are needed. This thesis addresses these challenges by combining semantic communication with generative models to optimize image compression and resource allocation in edge networks. Unlike traditional bit-centric communication systems, semantic communication prioritizes the transmission of meaningful data specifically selected to convey the meaning rather than obtain a faithful representation of the original data. The communication infrastructure can benefit of the focus solely on the relevant parts of the data due to significant improvements in bandwidth efficiency and latency reduction. Central to this work is the design of semantic-preserving image compression algorithms, utilizing advanced generative models such as Generative Adversarial Networks and Denoising Diffusion Probabilistic Models. These algorithms compress images by encoding only semantically relevant features and exploiting the generative power at the receiver side. This allows for the accurate reconstruction of high-quality images with minimal data transmission. The thesis also introduces a Goal-Oriented edge network optimization framework based on the Information Bottleneck problem and stochastic optimization, ensuring that communication resources are dynamically allocated to maximize efficiency and task performance. By integrating semantic communication into edge networks, the proposed system achieves a balance between computational efficiency and communication effectiveness, making it particularly suited for real-time applications. The thesis compares the performance of these semantic communication models with conventional image compression techniques, using both classical and semantic-awar

关键词： Image compression

来源：评论

学校读者我要写书评

暂无评论

Deep learning for hyperspectral image classification: A comprehensive review and future predictions

引用

Information Fusion 2025年 123卷

作者： Song, Yongchao Zhang, Junhao Liu, Zhaowei Xu, Yang Quan, Siwen Sun, Lijun Bi, Jiping Wang, Xuan School of Computer and Control Engineering Yantai University Yantai264005 China Science and Technology Office Yantai University Yantai264005 China School of Electronics and Control Engineering Chang'an University Xi'an710064 China

Hyperspectral image classification (HSIC) is an important research direction in the field of remote sensing image analysis and computer vision, which is of great practical significance. Hyperspectral imaging (HSI) is widely used in a variety of scenarios with its rich spectral and spatial information, but problems such as high-dimensional data characteristics and scarcity of labeled samples challenge the classification accuracy. Deep learning (DL), with its powerful feature extraction and modeling capabilities, provides an effective means to solve the nonlinear problems in HSIC. In this survey, we systematically review the research progress and applications of DL in HSIC. Firstly, we outline the importance of accurate classification, analyze the features of HSI and the challenges faced by DL in this area. Secondly, we introduce different feature representations of HSI and provide a comprehensive describe of the application of various DL models in HSIC. Meanwhile, we also explore DL methods that can effectively improve the classification performance in the case of insufficient training samples. Finally, we summarize the current research situation, and put forward the future development direction and suggestions. © 2025 Elsevier B.V.

关键词： Labeled data

来源：评论

学校读者我要写书评

暂无评论

Bayes and Biased Estimators Without Hyper-Parameter Estimation: Comparable Performance to the Empirical-Bayes-Based Regularized Estimator

引用

IEEE control Systems Letters 2025年 9卷 294-299页

作者： Yue Ju Bo Wahlberg Håkan Hjalmarsson Division of Decision and Control Systems School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Stockholm Sweden Division of Decision and Control Systems and the Competence Centre for Advanced BioProduction by Continuous Processing AdBIOPRO School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Stockholm Sweden

Regularized system identification has become a significant complement to more classical system identification. It has been numerically shown that kernel-based regularized estimators often perform better than the maximum likelihood estimator in terms of minimizing mean squared error (MSE). However, regularized estimators often require hyper-parameter estimation. This letter focuses on ridge regression and the regularized estimator by employing the empirical Bayes hyper-parameter estimator. We utilize the excess MSE to quantify the MSE difference between the empirical-Bayes-based regularized estimator and the maximum likelihood estimator for large sample sizes. We then exploit the excess MSE expressions to develop both a family of generalized Bayes estimators and a family of closed-form biased estimators. They have the same excess MSE as the empirical-Bayes-based regularized estimator but eliminate the need for hyper-parameter estimation. Moreover, we conduct numerical simulations to show that the performance of these new estimators is comparable to the empirical-Bayes-based regularized estimator, while computationally, they are more efficient.

关键词： Bayes methods Maximum likelihood estimation Covariance matrices Vectors System identification Noise measurement Linear regression Probability density function Contracts Computational efficiency

来源：评论

学校读者我要写书评

暂无评论

Energy-Free Sensing and Context Recognition Using Photovoltaic Cells

引用

IEEE Sensors Journal 2025年第9期25卷 15598-15611页

作者： Shintani, Kaede Rizk, Hamada Yamaguchi, Hirozumi Osaka University Graduate School of Information Science and Technology Department of Information Networking Suita565-0871 Japan Tanta University Tanta31733 Egypt RIKEN Center for Computational Science Kobe650-0047 Japan RIKEN Center for Computational Science Department of Computer and Control Engineering Kobe650-0047 Japan

The field of energy-free sensing and context recognition has recently gained significant attention as it allows operating systems without external power sources. Photovoltaic cells can convert light energy into electrical energy to power sensing devices, but their power may not be sufficient to ensure energy-free sensing due to the varying power needs of sensors and high computational demands. In this article, we propose the use of photovoltaic cells as a standalone sensor for the recognition of different contexts, including user identification, step counting, and location tracking. The system utilizes the photocurrent readings generated by the photovoltaic cells to capture the unique mobility patterns of different users. By analyzing these patterns, the system can accurately identify the user, count the number of steps taken, and track the user’s location. We propose a computationally efficient dynamic time warping (DTW) to match the variable length sequences of photocurrent readings to a database of known patterns and identify the closest subject and location matches. The system was rigorously evaluated in a realistic environment, and the results indicate that it can accurately estimate step count, identify subjects, and localize them with an accuracy of 88%, 90%, and 43 cm, respectively. This is achieved while the proposed system is nonintrusive and can operate without external power sources, making it a promising technology for energy-free sensing and context recognition. © 2001-2012 IEEE.

关键词： Photovoltaics

来源：评论

学校读者我要写书评

暂无评论

Optimization Scheduling of Multi-Energy Microgrid Based on the QUBO Quantum Computing Model 19th

Optimization Scheduling of Multi-Energy Microgrid Based on t...

引用

19th Annual Conference of China Electrotechnical Society, ACCES 2024

作者： Wang, Baonan Wang, Hui Zhang, Dan College of Computer Science and Technology Shanghai University of Electric Power Shanghai China State Key Laboratory of Power System Operation and Control Dept. of Electrical Engineering Tsinghua University Beijing China Institute of Logistic Science and Engineering Shanghai Maritime University Shanghai China

ISBN: (纸本)9789819608966

Current optimization methods for microgrid scheduling face issues such as insufficient precision in energy distribution, high operational costs, and inefficiency. In response to these challenges, an optimization scheduling method for multi-energy microgrids based on the QUBO (Quadratic Unconstrained Binary Optimization) quantum computing model is proposed in this paper, alongside a discretization approach based on a grid partitioning strategy. To validate the feasibility of addressing microgrid optimization scheduling problems in a quantum computing environment, a QUBO model is constructed and simulation studies are conducted in the qbsolv environment provided by D-Wave. The results indicate that operational costs are significantly reduced with an increase in the number of grids, particularly a reduction of 4.52% when the number of grids increases from five to ten. This method not only effectively enhances the precision of electricity supply and energy utilization efficiency but also reduces unnecessary electrical waste and economic costs. © Beijing Paike Culture Commu. Co., Ltd. 2025.

关键词： Microgrids

来源：评论

学校读者我要写书评

暂无评论

An Entropy-Integrated Adaptive Coding and Scheduling Framework for Optimized Data Transmission in Fog-Cloud IoT Architectures

引用

IEEE Internet of Things Journal 2025年

作者： Lu, Yang Zang, Yuting Bian, Ziyi Zhao, Shuai Zheng, Yan Northeast Forestry University College of Computer and Control Engineering Harbin China Harbin Normal University College of Computer and Information Engineering Harbin China La Trobe University School of Computing Engineering and Mathematical Sciences MelbourneVIC Australia

Entropy-driven network coding has been used as a basis for optimizing data transmission and enhancing resource utilization in Fog-Cloud IoT architectures, including applications in smart cities, industrial automation, environmental monitoring, and healthcare. In Fog-Cloud IoT architectures, conventional data transmission protocols are inefficient because they cannot adapt to dynamic entropy levels, resulting in underutilized bandwidth, increased latency, and higher energy consumption. In this paper, we propose an Enhanced Entropy-Driven Network Coding (E-EDNC) framework to address these problems. Our framework integrates real-time entropy estimation with adaptive coding strategies and employs a hybrid evolutionary-reinforcement learning (HE-RL) algorithm to dynamically optimize coding parameters and scheduling decisions. Experimental results demonstrate that E-EDNC improves bandwidth utilization by 25%, reduces latency by 25%, and decreases energy consumption by 12%, thereby enhancing overall data transmission efficiency and reliability in Fog-Cloud IoT environments. © 2014 IEEE.

关键词： Network coding

来源：评论

学校读者我要写书评

暂无评论

Streamlining the New Product Introduction by Integrating the Escalation Process 18th

Streamlining the New Product Introduction by Integrating the...

引用

18th International Conference on Interdisciplinarity in engineering, INTER-ENG 2024

作者： Zgirdea, Mihai Ionica, Andreea Cristina Leba, Monica Engineering and Management Doctoral School University of Petrosani 20 Universitatii Petrosani332006 Romania Management and Industrial Engineering Department University of Petrosani 20 Universitatii Petrosani332006 Romania Computer and System Control Engineering Department University of Petrosani 20 Universitatii Petrosani332006 Romania

ISBN: (纸本)9783031813771

In the dynamic landscape of modern business, companies face the challenge of rapidly introducing new products to meet consumer demands. This research aims to propel NPI efficiency, focusing on the automotive industry, by enhancing cooperative effectiveness and streamlining processes. The overarch- ing objective is to expedite high-quality product with the focus of introducing a controlled escalation process to facilitate swift corrections and improvements throughout each phase of NPI. To realize these objectives, a comprehensive NPI system has been developed. Integration of an escalation process into this sys- tem enables hierarchical escalation of concerns, blocking points, and new pro- cesses, addressing common project delays. The proposed software stands out for its unique integration of the escalation process, ensuring that all organizational members are promptly informed of any deviations using a dynamic project status feature, adapting as the project progresses. The integrated escalation process not only addresses anomalies promptly but also quantifies the time taken to restore normalcy or implement corrections, thereby fostering a more responsive and effi- cient NPI framework. The paper is organized into sections comprising a literature review on NPI methodologies, the integration of an escalation process into the NPI framework, a case study illustrating the implementation of the escalation process in an existing NPI methodology, and the subsequent results. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Automotive Industry Collaborative Teamwork controlled Escalation Process Dynamic NPI Software New Product Introduction (NPI)

来源：评论

学校读者我要写书评

暂无评论

Description and comparison of fault detection algorithms based on a selected building automation device

引用

Bulletin of the Polish Academy of Sciences: Technical Sciences 2025年第3期73卷

作者： Wlostowska, Sandra Kawa, Bartlomiej Borkowski, Piotr Department of Electrical Apparatus Faculty of Electrical Electronics Computer and Control Engineering Lodz University of Technology Lódź Poland

The article compares selected classification algorithms and those dedicated to anomaly detection. The models used temperature measurements in four rooms simulated in the MATLAB Simscape environment as test signals. The empirical part of the work consists of two parts. In the first one, an example data from the simulated building heating model object, models were built using unsupervised and supervised machine learning algorithms. Then, data from the facility was collected again with changed parameters (failures occurred at times other than the test ones, and the temperature patterns differed from those recorded and used to train the models). The algorithm effects and test signals (temperature changes) were saved in the database. The results were presented graphically in the Grafana program. The second part of the work presents a solution in which the analysis of the operating status of the heating system takes place in real time. Using an OPC server, data was exchanged between the MATLAB environment and the database installed on a virtual machine in the Ubuntu system. The conclusions present the results and collect the authors’ suggestions regarding the practical applications of the discussed classification models. © 2025 The Author(s).

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Quality-Relevant Modeling and Monitoring of Industrial Cyber-Physical Systems: The Semi-Supervised Dynamic Latent Variable Models

IEEE Transactions on Industrial Cyber-Physical Systems

引用

IEEE Transactions on Industrial Cyber-Physical Systems 2025年 3卷 39-47页

作者： Zhou, Le Wang, Yaoxin Wu, Yuanqing He, Shenghuang Song, Zhihuan Zhejiang University of Science and Technology Department of Automation and Electrical Engineering Hangzhou310023 China Guangdong University of Technology Department of Automation Guangzhou510006 China Dongguan University of Technology Department of Computer Science and Technology Dongguan523808 China Ningbo Industrial Internet Institute Ningbo315012 China Zhejiang University Department of Control Science and Engineering Hangzhou310027 China

In modern industrial cyber-physical systems, a mass of process variables has been obtained by the high-sampling online sensors. Meanwhile, the key quality indexes are usually obtained infrequently from the laboratory. Hence, these quality variables are with low sampling rate. To avail of the complete process and quality variables with various sampling rates in the dynamic processes, a set of semi-supervised dynamic latent variable models are proposed for dynamic modeling and quality-relevant monitoring. The proposed models have built a unified structure to consider both the auto-correlations and cross-correlations between the process and quality variables with unbalanced sampling sizes. Hence, the feature extraction of the time series data is dynamically adjusted under the guidance of the quality variables. Then, the quality-relevant monitoring schemes are proposed, which is validated by a numerical case and an actual wastewater treatment process. © 2024 IEEE. All rights reserved.

关键词： Indexes Probabilistic logic Kalman filters Noise Cyber-physical systems Analytical models Parameter estimation Time series analysis Process monitoring Numerical models

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：