检索结果-内蒙古大学图书馆

Driver Drowsiness Detection Using Swin Transformer and Diffusion Models for Robust Image Denoising

IEEE Access 2025年 13卷 71880-71907页

作者： El-Nabi, Samy Abd Ibrahim, Ahmed F. El-Rabaie, El-Sayed M. Hassan, Osama F. Soliman, Naglaa F. Ramadan, Khalil F. El-Shafai, Walid Faculty of Computer Science and Engineering Department of Artificial Intelligence Engineering South Sinai 46511 Egypt Menoufia University Faculty of Electronic Engineering Department of Electronics and Electrical Communications Engineering Menouf32952 Egypt Suez Canal University Faculty of Computers and Informatics Information Systems Department Ismailia41522 Egypt Princess Nourah bint Abdulrahman University College of Computer and Information Sciences Department of Information Technology P.O. Box 84428 Riyadh11671 Saudi Arabia Computer Science Department Riyadh11586 Saudi Arabia

With the rapid development of intelligent transportation systems and growing emphasis on driver safety, real-time detection of driver drowsiness has become a critical area of research. This study presents a robust and scalable driver drowsiness detection framework that integrates a Swin Transformer-based deep learning model with a diffusion model for image denoising. While conventional convolutional neural networks (CNNs) are effective in standard vision tasks, they often suffer performance degradation in real-world driving scenarios due to noise, poor lighting, motion blur, and adversarial attacks. To address these challenges, the proposed model focuses on eye-state detection, specifically, prolonged eye closure, as a primary indicator of driver disengagement and fatigue. Our system introduces a novel preprocessing stage using a denoising diffusion model built on a U-Net encoder-decoder architecture, effectively mitigating the impact of Gaussian noise and adversarial perturbations. Additionally, we incorporate adversarial training with Fast Gradient Sign Method (FGSM) and Projected Gradient Descent (PGD) attacks, demonstrating significant improvements in classification accuracy and resilience. Evaluations are conducted on two benchmark datasets, Eye-Blink and Closed Eyes in the Wild (CEW), under both clean and noisy conditions. Comparative experiments show that the proposed system outperforms several state-of-the-art models, including ViT, ResNet50V2, InceptionV3, MobileNet, DenseNet169, and VGG19, in terms of accuracy (up to 99.82%), PSNR (up to 41.61 dB), and SSIM (up to 0.984), while maintaining competitive inference times suitable for practical deployment. Moreover, a detailed sensitivity analysis of data augmentation strategies reveals that techniques such as rotation and horizontal flip substantially enhance the model’s generalization across variable visual inputs. The system also demonstrates improved robustness under real-world black-box scenarios and adver

关键词： Sensitivity analysis

来源：评论

学校读者我要写书评

暂无评论

A Novel Pressure Sensor with Composite Sensitive Film for High Precision Measurement at Broad Range 38

A Novel Pressure Sensor with Composite Sensitive Film for Hi...

引用

38th IEEE International Conference on Micro Electro Mechanical Systems, MEMS 2025

作者： Chen, Xiaopeng Ling, Sijia Tong, Hanyang Xiao, Yujing Zhang, Jin Yu, Zhengyin Jin, Qinghui Ningbo University The Faculty of Electrical Engineering and Computer Science China Chinese Academy of Sciences Shanghai Institute of Microsystem and Information Technology China

ISBN: (纸本)9798331508890

Microelectromechanical system (MEMS) based pressure sensors have been utilized for decades;however, new trends in pressure sensors have recently emerged, such as increased sensitivity, a broader range and reduced chip size. This paper presents a tiny 0.6mm×0.8mm piezoresistive pressure sensor that offers exceptional sensitivity (0.27mV/kPa/3.3V) at a wide range of pressures (0-1MPa). The innovative composite pressure-sensitive film of the sensor is composed of a 10μm "*"-shaped single-crystalline silicon structure and a remarkably thin thickness of only 2.5μm poly-silicon film. It is developed utilizing the thin-film under bulk-silicon technique. And simultaneous improvement of detection sensitivity and linearity is achieved by etching out the nonlinear deformation region of the film. © 2025 IEEE.

关键词： Pressure sensors

来源：评论

学校读者我要写书评

暂无评论

Experience in a VR Cricket Game Using Brain computer Interaction (BCI) 7

Experience in a VR Cricket Game Using Brain Computer Interac...

引用

7th IEEE International Conference on Artificial Intelligence and eXtended and Virtual Reality, AIxVR 2025

作者： Jayaraj, Lionel Wood, James Reifer, Amanda Braund, Edward School of Computer Science and Technology University of Bedfordshire Luton United Kingdom University of the West Indies Faculty of Sport Cave Hill Barbados

ISBN: (纸本)9798331521578

This vision paper outlines the necessity to improve the user-experience of Extended Reality (XR) simulations. XR applications use the term 'immersion' to describe the technological factors that lead to an effective user experience and improve user presence [4] [9]. This research aims to improve immersion by user's motion tracking for playing [1] [2] enhanced by EEG signals from Brain especially for the extreme cricket bowling activities. This requires many nuanced controls including motion control. Hence, this empirical research enables mapping of the natural user interface (NUI) in game play [7] along with the Brain computer Interaction (BCI) [8] to improve immersion. The research application development aims to pipeline a BCI device called 'NeuroConcise' [6] to acquire decision data for additional inputs to the commercial Virtual Reality (VR) Game controls from Meta [3]. © 2025 IEEE.

关键词： Cricket (Sports)

来源：评论

学校读者我要写书评

暂无评论

Intelligent wearable vision systems for the visually impaired in Saudi Arabia

引用

Neural Computing and Applications 2025年第14期37卷 8571-8591页

作者： Talaat, Fatma M. El-Shafai, Walid Soliman, Naglaa F. Algarni, Abeer D. El-Samie, Fathi E. Abd Faculty of Artificial Intelligence Kafrelsheikh University Kafrelsheikh33516 Egypt Faculty of Computer Science and Engineering New Mansoura University Gamasa35712 Egypt Security Engineering Lab Computer Science Department Prince Sultan University Riyadh11586 Saudi Arabia Department Electronics and Electrical Communications Engineering Faculty of Electronic Engineering Menoufia University Menouf32952 Egypt Department of Information Technology College of Computer and Information Sciences Princess Nourah Bint Abdulrahman University P.O. Box 84428 Riyadh11671 Saudi Arabia

Navigating the world with visual impairments presents unique challenges, often limiting independence and safety. This research introduces SafeStride, a novel algorithm designed to empower visually impaired individuals through real-time obstacle detection and navigation assistance. SafeStride functions through five interconnected phases: sensor data acquisition, data preprocessing, obstacle detection, path planning, and feedback and guidance. A combination of ultrasonic sensors, cameras, and inertial measurement units (IMUs) feed comprehensive environmental data into the system, enabling accurate obstacle detection and safe path planning. Evaluating SafeStride's performance involved the MNIST dataset for image recognition and a dedicated indoor object detection dataset. Additionally, comparisons were made against three deep learning models—convolutional neural network (CNN), long short-term memory (LSTM), and gated recurrent unit (GRU). The results showcased SafeStride's high accuracy and effectiveness in both obstacle detection and classification. Beyond its impressive performance, SafeStride represents a significant leap forward in navigation aids for the visually impaired. By offering a comprehensive solution that enhances safety and independence in diverse environments, it holds immense potential to improve the lives of countless individuals. Future efforts will focus on further optimizing the algorithm and testing its real-world capabilities, paving the way for a future where safe and independent navigation becomes a reality for all. All three deep learning models (CNN, LSTM, GRU) excelled in the experiment, achieving high accuracy above 0.98, strong recall and F1-scores near 0.99, and exceptional AUC-ROC scores exceeding 0.9998. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2025.

关键词： Motion planning

来源：评论

学校读者我要写书评

暂无评论

A Hybrid Embedding Approach to Classify Crisis-Related Short Text 12th

A Hybrid Embedding Approach to Classify Crisis-Related Short...

引用

12th International Conference on Recent Trends in Computing, ICRTC 2024

作者： Nath, Basudev Sahoo, Deepak Paul, Nayan Ranjan Faculty of Engineering Technologies Sri Sri University Cuttack India Department of Computer Science and Engineering Silicon Institute Technology Bhubaneswar India

ISBN: (纸本)9789819788354

High mobile phone usage and internet access on phones make it simple to connect to the globe and share your thoughts, feelings, and views on local or global issues on social media. This social media content helps government and corporate institutions grasp public opinion. People utilise social media to report disasters and request medical, food, and water aid. It is difficult to recognise disaster-related tweets and determine people's needs. This paper proposed a hybrid model to recognise disaster-related tweets. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Tweets

来源：评论

学校读者我要写书评

暂无评论

TinyML Model for Solar Cell Defect Classification Based on Electroluminescence Images 4th

TinyML Model for Solar Cell Defect Classification Based on E...

引用

4th International Conference on Electronic Engineering and Renewable Energy Systems, ICEERE 2024

作者： Boubaker, S. Moussaoui, C. Mellit, Adel Benghanem, M. Department of Computer and Network Engineering College of Computer Science and Engineering University of Jeddah Jeddah21959 Saudi Arabia Faculty of Sciences and Technology University of Jijel Jijel18000 Algeria Department of Physics Faculty of Science Islamic University of Madinah Madinah42351 Saudi Arabia

ISBN: (纸本)9789819606436

Being largely utilized as alternative sources of energy, photovoltaic (PV) panels are being deployed in many locations. However, those panels are extensively being prone to faults. A careful detection and diagnosis of those faults becomes a must for solar energy engineers, practitioners and decision-makers in order to extract the maximum of energy from them. The integration of artificial intelligence (AI) in the domain of solar energy management has become a trend in the few previous years because of the high potential provided by machine learning (ML) and deep learning (DL) algorithms in handling issues related to PV systems such as control and fault detection and isolation (FDI). In this paper, TinyML models for defect classification of photovoltaic modules based on electroluminescence (EL) images were developed. An open-source platform, Edge Impulse, was used to design and deploy the models. Two convolutional neural network models (mobileNetV1 and mobileNetV2) have been used and compared. A dataset of EL with three anomalies was used to develop the models. The validation test shows an accuracy of 88.44 and 76.38% for respectively MobileNet-V1 and MobileNet-V2. The quantized (Int8) model has been integrated into a microcontroller (Arduino Portenta H7) for real-time application. The obtained results can be considered as promising and can be extended to other case studies with more defects and larger datasets. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

Mapping the Scientific Landscape of Integrated Scheduling Towards Industry 5.0 with A Bibliometric Analysis 6

Mapping the Scientific Landscape of Integrated Scheduling To...

引用

6th International Conference on Industry 4.0 and Smart Manufacturing, ISM 2024

作者： Putri Ardyanti, Anak Agung Ayu Xie, Zhiqiang Karnyoto, Andrea Stevens Pardamean, Bens Estiyanti, Ni Made School of Computer Science and Technology Harbin University of Science and Technology Harbin150000 China Faculty of Information Technology and Design Primakara University Bali 80226 Indonesia Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta11480 Indonesia Computer Science Department BINUS Graduate Program Master of Computer Science Bina Nusantara University Jakarta 11480 Indonesia School of Management and Business University of Debrecen Debrecen Hungary

The industry is rapidly transitioning from the 4.0 era to the 5.0 era, prompting renewed interest among scholars in scheduling problems. They allow operations to process and assemble various components simultaneously. This bibliometric analysis provides a comprehensive understanding of the diverse research perspectives on the potential challenges of integrated scheduling within the framework. We identified 357 Scopus articles and analysed co-occurrence keyword (CNK) by VOSViewer tools and Global Citation Score (GCS) by Scopus tools. The research data obtained from 1993 to 2023 mainly in the subject areas of Engineering, computer science, Maths, and others. CNK identified six cluster keywords, and nine most cited articles based on normalised GCS. These results contribute to our understanding of the development of scheduling algorithms, shifting from an emphasis on manufacturing transfers to contemporary demand-centric research directions. Furthermore, this research provides a valuable perspective for policymakers, plant personnel, and manufacturing managers to make informed decisions. © 2025 The Authors. Published by Elsevier B.V.

关键词： Job shop scheduling

来源：评论

学校读者我要写书评

暂无评论

Secure and Optimized Cloud Storage with Privacy Preservation using Improved SHA - 256 2

Secure and Optimized Cloud Storage with Privacy Preservation...

引用

2nd International Conference on Computational Intelligence, Communication technology and Networking, CICTN 2025

作者： Joshi, Aditi Devarajan, Ganesh Gopal Karuppiah, Marimuthu Sharma, Yogesh Kumar Department of Computer Science and Engineering SRM Institute of Science and Technology Delhi NCR Campus Uttar Pradesh Ghaziabad201204 India Department of Computer Science and Engineering Faculty of Engineering and Technology SRM Institute of Science and Technology Delhi NCR Campus Uttar Pradesh Ghaziabad201204 India Presidency School of Computer Science and Engineering Presidency University Karnataka Bengaluru560064 India Department of CSE Dewan V S Institute of Technology U.P Meerut India

ISBN: (纸本)9798331530389

With the increased reliance of enterprises on cloud storage solutions for data management, maintaining safe and efficient storage while emphasizing privacy protection is crucial. With an emphasis on user privacy, this study examines the methods, tools, and difficulties currently involved in attaining safe and effective cloud storage. It looks at access control protocols, encryption techniques like SHA-256, and new technologies that are intended to improve the security and effectiveness of cloud storage systems. The study also addresses the changing trends influencing the security of cloud storage in the future and emphasizes the significance of adhering to privacy laws. © 2025 IEEE.

关键词： Cloud storage

来源：评论

学校读者我要写书评

暂无评论

MARIE:One-Stage Object Detection Mechanism for Real-Time Identifying of Firearms

引用

computer Modeling in Engineering & sciences 2025年第1期142卷 279-298页

作者： Diana Abi-Nader Hassan Harb Ali Jaber Ali Mansour Christophe Osswald Nour Mostafa Chamseddine Zaki Computer Science Department Faculty of SciencesLebanese UniversityBeirut146404Lebanon College of Engineering and Technology American University of the Middle EastEgaila54200Kuwait Lab-STICC CNRS UMR 6285ENSTA-BretagneBrest29200France

Security and safety remain paramount concerns for both governments and individuals *** today’s context,the frequency of crimes and terrorist attacks is alarmingly increasing,becoming increasingly intolerable to ***,there is a pressing need for swift identification of potential threats to preemptively alert law enforcement and security forces,thereby preventing potential attacks or violent *** advancements in big data analytics and deep learning have significantly enhanced the capabilities of computer vision in object detection,particularly in identifying *** paper introduces a novel automatic firearm detection surveillance system,utilizing a one-stage detection approach named MARIE(Mechanism for Realtime Identification of Firearms).MARIE incorporates the Single Shot Multibox Detector(SSD)model,which has been specifically optimized to balance the speed-accuracy trade-off critical in firearm detection *** SSD model was further refined by integrating MobileNetV2 and InceptionV2 architectures for superior feature extraction *** experimental results demonstrate that this modified SSD configuration provides highly satisfactory performance,surpassing existing methods trained on the same dataset in terms of the critical speedaccuracy *** these innovations,MARIE sets a new standard in surveillance technology,offering a robust solution to enhance public safety effectively.

关键词： Firearm and gun detection single shot multi-box detector deep learning one-stage detector MobileNet inception convolutional neural network

来源：评论

学校读者我要写书评

暂无评论

A Data science Approach for Predicting Soccer Passes Using Positional Data 26th

A Data Science Approach for Predicting Soccer Passes Using ...

引用

26th International Conference on Information Integration and Web Intelligence, iiWAS 2024

作者： Eigenrauch, Sebastian Bischofberger, Jonas Baca, Arnold Schikuta, Erich Research Group Workflow Systems and Technology Faculty of Computer Science University of Vienna Vienna Austria Centre for Sport Science and University Sports University of Vienna Vienna Austria

ISBN: (纸本)9783031780899

Data-driven approaches for evaluating tactical team behavior in soccer are nowadays a widespread method in sport analytics. The large amount of data collections enables experts to generate a deep tactical understanding and extract valuable measurements out of team-performances. However, these approaches are often limited in their comprehensibility and applicability for domain experts. Additionally, defensive behaviour in soccer is notoriously difficult to measure and has been receiving less attention in research and practice compared to measuring offensive performance. The motivation of this research is the design, implementation and validation of data science algorithms, that predict tactical motion of defending players after an occurring event of a pass, one of the most common events in soccer matches. The focus is the establishment and validation of different sets of rules, which simulate the movement behavior of the defending team, based on domain knowledge. The approach provides a high level of applicability for domain experts, in order to use and combine variable predefined rules for prediction, simulation and evaluation of different tactical approaches of defensive behavior. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Data science

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：