检索结果-内蒙古大学图书馆

Meta-analysis of quantum convolutional neural networks for automated tuberculosis screening on chest x-rays

Multimedia Tools and Applications 2025年 1-27页

作者： Rana, Anurag Bhaglani, Dimple Kumar Vaidya, Pankaj Hu, Yu-Chen Computers and Data Sciences Yogananda School of AI Shoolini University Himachal Pradesh Solan India Department of Pulmonary Medicine Indira Gandhi Medical College Himachal Pradesh Shimla India Department of Computer Science Tunghai University No. 1727 Sec. 4 Boulevard Xitun District Taichung City407224 Taiwan

The problem of detecting tuberculosis (TB) from chest X-ray (CXR) images was addressed by developing a convolutional neural network (CNN) model. The CNN was trained using two publicly available chest radiograph datasets from Himachal Pradesh, India. To evaluate its performance, the CNN model was compared with a transfer learning-based technique that utilized several pre-trained CNNs. The results showed that the CNN model outperformed the transfer learning approach. The study further explored the integration of quantum algorithms in image classification. It introduced a novel hybrid quantum–classical computing paradigm, wherein classically challenging elements of an algorithm are handled by a quantum computer. Specifically, a hybrid quantum–classical image classification technique, inspired by CNNs and called the quanvolutional neural network (QNN), was used for diagnosing TB from chest X-rays. This new QNN variant employed novel enhanced quantum representation (NEQR) image encoding to transform pixel values into quantum states. The QNN model was trained on the same two chest radiograph datasets as the CNN model. The QNN demonstrated superior performance to the CNN model, with a validation accuracy of 87% during training. These findings suggest that hybrid quantum–classical image classification algorithms can surpass traditional methods in diagnosing tuberculosis from chest X-rays. From the results, the Exception, ResNet50, and VGG16 models performed exceptionally well in Quantum CNN models with image augmentation, achieving over 89% accuracy, precision, sensitivity, and F1-score for TB detection. The study suggests that a larger dataset could improve model precision and reliability. Differences in outcomes compared to previous studies may stem from using simulators of flawless quantum computers in earlier research. This study highlights the potential of integrating a quantum circuit with classical CNNs for TB diagnosis from chest X-rays using NEQR image encoding, pr

关键词： Quantum computers

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Evaluation of Machine Learning Algorithms for Harvested Arecanut Variety Classification and Detection

引用

Procedia computer science 2025年 258卷 4147-4156页

作者： Harsha Keladi Ganapathi Md Moinuddin Bhuiyan University of New Haven Master of Science in Data Science West Haven CT 06516 USA University of New Haven Department of Electrical and Computer Science West Haven CT 06516 USA

Traditional image processing techniques for nut classification often struggle due to the subtle visual differences between nut varieties, natural variations within varieties, and complexities in isolating nuts from cluttered backgrounds. This study explores the proficiency of machine learning and deep learning methods for accurate and efficient classification and detection of arecanut varieties. The Roboflow tool was used to meticulously label and annotate a dataset of 9,000 images, categorizing them into Isolated Green Arecanut, Isolated Yellow Arecanut, Clustered Green Arecanut, and Clustered Yellow Arecanut. The dataset also includes varied real-life scenarios to ensure robust validations of the model. Our classification process used classifiers like Support Vector Machines, K-nearest neighbor, and Random Forest. Object detection was performed using YOLOv8 and YOLO-NAS architectures. The study involved preprocessing, feature extraction (for classification), model training, and rigorous evaluation using metrics like accuracy, precision, recall, and mean average precision (mAP). This comparative analysis provides insights into the strengths and weaknesses of each algorithm, paving the way for robust automated systems in the agriculture industry. Among the four different classifiers, SVM yields the highest accuracy of 88%, highlighting its effectiveness in the classification task. Among the two detection algorithms, YOLOv8 has achieved 79.5% mean average precision and YOLO-NAS has achieved 78.5% mean average precision, while the best precision was produced by YOLO-NAS (82.5%).

关键词： YOLOv8 YOLO NAS SVM KNN Random Forest HOG Arecanut Mean Average Precision (mAP)

来源：评论

学校读者我要写书评

暂无评论

Formula One: Racing Intelligence in the Age of AI

Formula One: Racing Intelligence in the Age of AI

引用

Advancements in Electrical, Electronics, Communication, Computing and Automation (ICAECA), International Conference on

作者： R.K. Kavitha S. Kanishk Department of Computer Applications Kumaraguru College of Technology Coimbatore Department of Data Science Kumaraguru College of Liberals Arts and Science Coimbatore

ISBN: (数字)9798331525439

ISBN: (纸本)9798331525446

This study proposes an all-inclusive, data-driven system to optimize the performance of Formula One racing by incorporating state-of-the-art technology such as blockchain, edge computing, artificial intelligence, and real-time telemetry. The current systems of Formula 1 depend heavily on manual analysis and traditional data collection, thereby limiting the ability to make decisions in real time. The proposed system, however, incorporates high-resolution telemetry sensors with Nvidia Jetson edge computing devices, AI-driven prediction models, and secure blockchain protocols so that performance improvement during a race can be achieved even faster and more accurately. The key components are in the form of telemetry sensors for real-time data acquisition, edge computing for real-time processing, blockchain technology for secure data exchange and artificial intelligence models for predictive analytics. This method improves vehicle performance, adjusts to the race condition changes, and enhances general decision-making with an efficient, transparent, and scalable solution. The system architecture also accommodates future technological development as well as cooperation within the motorsport industry. The implications of the findings spread wider to driverless car industries and other data-dependent sectors, showing how blockchain and AI can change the game plan of racing and corporate formations.

关键词： Computational modeling Predictive models Real-time systems Sensor systems data models Blockchains Telemetry Artificial intelligence Intelligent sensors Edge computing

来源：评论

学校读者我要写书评

暂无评论

Identifying Sources of Soil Pollution in Industrial Areas Using IoT data and Neural Networks

Identifying Sources of Soil Pollution in Industrial Areas Us...

引用

2024 Asian Conference on Intelligent Technologies, ACOIT 2024

作者： Kovarasan, Rajesh Kambattan Shibu, N.V. Sairam, A. Manivannan, K.K. Ishwarya, M.V. GaneshBabu, T.R. Department of Computer Science and Engineering Vel Tech Rangarajan Dr.Sagunthala R&D Institute of Science and Technology Tamil Nadu Chennai India Department of Computing Technologies School of Computing SRM Institute of Science and Technology Kattankulathur Tamil Nadu Chennai India Department of Computer Science and Engineering Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Tamil Nadu Chennai India Department of Mechatronics Engineering KCG College of Technology Tamil Nadu Chennai India Artificial Intelligence and Data science Department Agni College of Technology Tamil Nadu Chennai India Tamil Nadu Namakkal India

ISBN: (纸本)9798350374933

Quick and precise source identification is crucial for developing successful remediation solutions to address soil contamination in industrial locations, which presents serious dangers to both the environment and human health. To locate the origins of soil contamination, this research presents a new method that combines neural network models with data collected from the Internet of Things (IoT). To monitor soil type, pollution levels, and weather conditions in real-time, set up IoT devices. After that, advanced neural network algorithms are used to process and analyze the data to find connections and patterns that might indicate the origins of the pollution. Our findings show that this method is useful for identifying where soil pollution inside industrial zones is coming from. Achieve high accuracy in source identification by integrating IoT technology and machine learning approaches. This allows us to address environmental harm and protect public health via focused remediation activities. Policymakers, environmental scientists, and stakeholders engaged in sustainable development and pollution control may benefit greatly which highlight the potential of IoT data and neural networks in challenges difficult environmental concerns. © 2024 IEEE.

关键词： Environmental monitoring

来源：评论

学校读者我要写书评

暂无评论

Liver Tumor Detection Using YOLOv8: Size-Adaptive Nanoparticle Design Based on Tumor Dimensions

Liver Tumor Detection Using YOLOv8: Size-Adaptive Nanopartic...

引用

Computational Intelligence Approaches and Applications (ICCIAA), International Conference on

作者： Laith Al Shehab Rehab Abdulrahman Alsaidi Rana Haitham Alraqad Department of Computer Science Middle East University Amman Jordan Department of Computer Vision Optriment Computer Systems Amman Jordan Department of Data Science App Trainers Amman Jordan

ISBN: (数字)9798331523657

ISBN: (纸本)9798331523664

Today, liver tumors have become a crucial and popular problem in our current era. Advances in artificial intelligence, especially computer vision, have enhanced diagnostic capabilities and provided deeper insights into tumor features. In this study, we leverage YOLOv8s for liver tumor detection, using MRI images to identify tumors with bounding box annotations and extract critical features such as shape, size, and diameter. These features are then used to determine the optimal nanoparticle size based on predefined rules, ensuring a tailored approach to treatment. Our framework demon- strates how AI can bridge the gap between tumor detection and nanoparticle design. The study revealed that tumor diameter significantly influences the optimal nanoparticle design. For tumor diameters of 257.99 µm, 297.47 µm, and 422.3 µm, the corresponding nanoparticle sizes were 729.0 nm, 535.0 nm, and 447.0 nm, respectively. Experimental results confirm the detection system's accuracy and effectiveness in recommending nanoparticle sizes, aiding liver tumor diagnosis and treatment.

关键词： Nanoparticles computer vision Liver cancer Targeted drug delivery Shape Magnetic resonance imaging Liver Feature extraction Artificial intelligence Tumors

来源：评论

学校读者我要写书评

暂无评论

Channel-wise attention based binarization for CNN Architectures 5

Channel-wise attention based binarization for CNN Architectu...

引用

5th International Conference on Emerging Systems and Intelligent Computing, ESIC 2025

作者： Paul, Ipsita Bandyopadhyay, Mainak Dash, Satya Ranjan Aurpa, Tanjim Taharat Kiit D.U. School of Computer Engineering Bhubaneswar India Kiit D.U. School of Computer Applications Bhubaneswar India Bsmr Digital University Department of Data Science Engineering Bangladesh

ISBN: (纸本)9798331522100

Attention mechanisms are known to effectively increase the performance of a model by focusing on specific sections of the input while executing a task. Squeeze and Excitation module (SE-Net) and Efficient channel attention module (ECA-Net) are well-known Channel Attention networks that only involve a few parameters to achieve a clear performance gain. In order to further improve the model's accuracy, Binarized Neural Networks (BNNs) can be introduced in the same. BNNs are deep learning models that use binarized values for activation functions and training weights, instead of fully precised values that leads to much faster computation, occupying less memory and power. Motivated by this, a method of binarization of the convolutional network is proposed in this paper that can enhance the overall model accuracy. In our work, we focused on evaluating the performance of our binarized attention module over the pre-existing ones. Thus, the objective of this work is to design, implement, and evaluate efficient neural network models that balance accuracy and computational efficiency for image classification tasks on datasets with differing complexity levels, such as CIFAR-10 and CIFAR-100. By exploring various backbone architectures (e.g., MobileNetV2 and ResNet50), attention mechanisms (e.g., ECANet and SENet), and binarization techniques, this study aims to understand the impact of these components on model performance and resource requirements. Ultimately, the goal is to develop optimized models that achieve high accuracy with minimal computational overhead, making them suitable for deployment in resource-constrained environments. © 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

EIGENVALUE BOUNDS FOR THE QUANTUM CHROMATIC NUMBER OF GRAPH POWERS

arXiv

引用

arXiv 2025年

作者： Abiad, Aida Jany, Benjamin Department of Mathematics and Computer Science Eindhoven University of Technology Netherlands Department of Mathematics and Data Science Vrije Universiteit Brussel Belgium

The quantum chromatic number, a generalization of the chromatic number, was first defined in relation to the non-local quantum coloring game. We generalize the former by defining the quantum k-distance chromatic number χkq(G) of a graph G, which can be seen as the quantum chromatic number of the k-th power graph, Gk, and as generalization of the classical k-distance chromatic number χk(G) of a graph. It can easily be shown that χkq(G) ≤ χk(G). In this paper, we strengthen three classical eigenvalue bounds for the k-distance chromatic number by showing they also hold for the quantum counterpart of this parameter. This shows that several bounds by Elphick et al. [J. Combinatorial Theory Ser. A 168, 2019, Electron. J. Comb. 27(4), 2020] hold in the more general setting of distance-k colorings. As a consequence we obtain several graph classes for which χkq(G) = χk(G), thus increasing the number of graphs for which the quantum parameter is known. © 2025, CC BY.

关键词： Eigenvalues and eigenfunctions

来源：评论

学校读者我要写书评

暂无评论

Enhancing Brain Tumor Detection with YOLOv11: An Innovative Deep Learning Approach in Medical Imaging

Enhancing Brain Tumor Detection with YOLOv11: An Innovative ...

引用

Computational Intelligence Approaches and Applications (ICCIAA), International Conference on

作者： Huthaifa Ziad Abuhammad Rehab Abdulrahman ALsaidi Rana Haitham Alraqad Department of Computer Science Middle East University Amman Jordan Department of Computer Vision OPTRIMENT COMPUTER SYSTEMS Amman Jordan Department of Data Science App Trainers Amman Jordan

ISBN: (数字)9798331523657

ISBN: (纸本)9798331523664

This study presents the application of YOLOv11 for the detection of brain tumors in MRI scans, achieving groundbreaking performance with a mean Average Precision at 50% intersection over union (mAP50) of 0.993, a precision rate of 98.8%, and a recall rate of 98.2%. These results significantly surpass the performance of previous state-of-the-art models in the field. The proposed YOLOv11 architecture integrates advanced techniques such as multiscale feature fusion, channel attention mechanisms, and an optimized loss function, which collectively contribute to its enhanced ability to detect small and subtle tumors. This innovation not only improves the accuracy of brain tumor detection but also reduces the time required for manual interpretation, which is crucial in clinical settings. The results demonstrate that YOLOv11 can serve as a reliable tool for enhancing diagnostic efficiency and ensuring more consistent detection outcomes, thus improving the overall reliability of automated systems in the medical imaging domain.

关键词： Deep learning Technological innovation Accuracy Magnetic resonance imaging Brain tumors Medical treatment Brain modeling Real-time systems Reliability Biomedical imaging

来源：评论

学校读者我要写书评

暂无评论

Privacy-Preserving Cyberattack Detection in Blockchain-Based IoT Systems Using AI and Homomorphic Encryption

引用

IEEE Internet of Things Journal 2025年第11期12卷 16478-16492页

作者： Manh, Bui Duc Nguyen, Chi-Hieu Hoang, Dinh Thai Nguyen, Diep N. Zeng, Ming Pham, Quoc-Viet University of Technology Sydney School of Electrical and Data Engineering SydneyNSW2007 Australia Université Laval Department of Electrical Engineering and Computer Engineering QuebecQCG1V 0A6 Canada University of Dublin School of Computer Science and Statistics Trinity College Dublin D02 PN40 Ireland

This work proposes a novel privacy-preserving cyberattack detection framework for blockchain-based Internet-of-Things (IoT) systems. In our approach, artificial intelligence (AI)-driven detection modules are strategically deployed at blockchain nodes to identify real-time attacks, ensuring high accuracy and minimal delay. To achieve this efficiency, the model training is conducted by a cloud service provider (CSP). Accordingly, blockchain nodes send their data to the CSP for training, but to safeguard privacy, the data is encrypted using homomorphic encryption (HE) before transmission. This encryption method allows the CSP to perform computations directly on encrypted data without the need for decryption, preserving data privacy throughout the learning process. To handle the substantial volume of encrypted data, we introduce an innovative packing algorithm in a Single-Instruction-Multiple-data (SIMD) manner, enabling efficient training on HE-encrypted data. Building on this, we develop a novel deep neural network training algorithm optimized for encrypted data. We further propose a privacy-preserving distributed learning approach based on the FedAvg algorithm, which parallelizes the training across multiple workers, significantly improving computation time. Upon completion, the CSP distributes the trained model to the blockchain nodes, enabling them to perform real-time, privacy-preserved detection. Our simulation results demonstrate that our proposed method can not only mitigate the training time but also achieve detection accuracy that is approximately identical to the approach without encryption, with a gap of around 0.01%. Additionally, our real implementations on various blockchain consensus algorithms and hardware configurations show that our proposed framework can also be effectively adapted to real-world systems. © 2025 IEEE.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

A Blockchain-Based Secure and Transparent Global Radioactive Source Lifecycle Management System

A Blockchain-Based Secure and Transparent Global Radioactive...

引用

IEEE Southeastcon

作者： Susan Poudel Firdous Kausar Debashis Das Pushpita Chatterjee Sajid Hussain Uttam Ghosh Dept. of Mathematics and Computer Science Fisk University Nashville TN USA Department of Computer Science and Data Science Meharry Medical College Nashville TN USA

ISBN: (数字)9798331504847

ISBN: (纸本)9798331504854

The management of radioactive sources is a critical process that ensures the safe and responsible handling of radioactive materials throughout their lifecycle. These sources require careful management from production to disposal, such as real-time tracking and radiation monitoring to follow everyone's safety rules and protect people and the environment. However, these sources present significant global challenges, especially regarding safety, security, and transparency. Current systems face limitations such as fragmented oversight, lack of accountability, and risks of unauthorized access. To address these limitations, this paper proposes a blockchain-based radioactive source lifecycle management system to manage the lifecycle of radioactive sources. Blockchain's decentralized and tamper-proof ledger secures data throughout the entire lifecycle of radioactive materials. By using smart contracts and access controls, the system ensures that only authorized parties can monitor and verify transactions in real-time, which reduces human error and prevents unauthorized changes to the data. Users can perform key operations such as retrieving source details, transferring ownership, updating source locations, and adding observers to our proposed system. Our experiment in designing and testing a blockchain application has proven the potential for a secure and transparent system that enhances global cooperation in managing radioactive sources. Overall, the proposed system not only addresses current challenges in radioactive source management but also enhances the tracking and monitoring of radioactive materials.

关键词： Access control Radioactive materials Smart contracts Supply chains Production Position measurement Real-time systems Blockchains Safety Testing

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：