检索结果-内蒙古大学图书馆

A Lightweight CNN Based on Transfer Learning for COVID-19 Diagnosis

computers, Materials & Continua 2022年第7期72卷 1123-1137页

作者： Xiaorui Zhang Jie Zhou Wei Sun Sunil Kumar Jha Wuxi Research Institute Nanjing University of Information Science&TechnologyWuxi214100China Engineering Research Center of Digital Forensics Ministry of EducationJiangsu Engineering Center of Network MonitoringSchool of Computer and SoftwareNanjing University of Information Science&TechnologyNanjing210044China Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology(CICAEET) Nanjing University of Information Science&TechnologyNanjing210044China School of Automation Nanjing University of Information Science&TechnologyNanjing 210044 China IT Fundamentals and Education Technologies Applications University of Information Technology and Management in RzeszowRzeszow Voivodeship100031Poland

The key to preventing the COVID-19 is to diagnose patients quickly and *** have shown that using Convolutional Neural Networks(CNN)to analyze chest Computed Tomography(CT)images is helpful for timely COVID-19 ***,personal privacy issues,public chest CT data sets are relatively few,which has limited CNN’s application to COVID-19 ***,many CNNs have complex structures and massive *** if equipped with the dedicated Graphics Processing Unit(GPU)for acceleration,it still takes a long time,which is not conductive to widespread *** solve above problems,this paper proposes a lightweight CNN classification model based on transfer *** the lightweight CNN MobileNetV2 as the backbone of the model to solve the shortage of hardware resources and computing *** order to alleviate the problem of model overfitting caused by insufficient data set,transfer learning is used to train the *** study first exploits the weight parameters trained on the ImageNet database to initialize the MobileNetV2 network,and then retrain the model based on the CT image data set provided by *** results on a computer equipped only with the Central Processing Unit(CPU)show that it consumes only 1.06 s on average to diagnose a chest CT *** to other lightweight models,the proposed model has a higher classification accuracy and reliability while having a lightweight architecture and few parameters,which can be easily applied to computers without GPU ***:***/ZhouJie-520/paper-codes.

关键词： Convolutional neural networks chest computed tomography image COVID-19 transfer learning mobileNetv2

来源：评论

学校读者我要写书评

暂无评论

A New Rhythm in A1: Convolutional Neural Networks for Music Genre Classification 18

A New Rhythm in A1: Convolutional Neural Networks for Music ...

引用

18th IEEE International Conference on Emerging Technologies, ICET 2023

作者： Javed, Huzaifa Arshad, Usama Abideen, Zain Ul Ali, Raja Hashim BCMaterials Basque Center for Materials Applications & Nanostructure Basque Country Leioa Spain Gik Institute of Engg. Sciences & Technology Faculty of Computer Science & Engineering Topi Pakistan University of Europe for Applied Sciences Dept. of Technology & Software Engineering Berlin Germany

ISBN: (纸本)9798350328172

Despite the significance of music genre classification in audio identification, it remains under-explored within AI research. This tool is crucial for personalized music recommendations and similar music detection. We have developed an efficient AI model that leverages Convolutional Neural Networks (CNNs), offering high-precision genre identification when integrated into a graphical user interface. The model effectively extracts audio features like Mel Frequency Cepstral Coefficients (MFCCs), zero-crossing rate, and tempo. Testing results reveal strong performance in genre prediction across diverse tracks, affirming the model's ability to discern unique characteristics of various music genres. This performance not only attests to the model's capability in discerning the unique characteristics inherent to different music genres but also suggests that it can effectively generalize to novel, unseen data. Our model lays the groundwork for future enhancements and demonstrates the potential of AI in transforming the music industry - from personalized music playlists to exploratory recommendation systems. The success of this model paves the way for more intricate applications of AI within music analysis. © 2023 IEEE.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

The YOLOv8 Edge: Harnessing Custom Datasets for Superior Real-Time Detection 18

The YOLOv8 Edge: Harnessing Custom Datasets for Superior Rea...

引用

18th IEEE International Conference on Emerging Technologies, ICET 2023

作者： Ahmed, Tafreed Maaz, Ahmad Mahmood, Danyaal Abideen, Zain Ul Arshad, Usama Ali, Raja Hashim BCMaterials Basque Center for Materials Applications & Nanostructure Basque Country Leioa Spain Gik Institute of Engg. Sciences & Technology Faculty of Computer Science & Engineering Topi Pakistan University of Europe for Applied Sciences Dept. of Technology & Software Engineering Berlin Germany

ISBN: (纸本)9798350328172

YOLO has become a central real-time object detection system for robotics, driverless cars, and video monitoring applications. Developing a custom object detection solution that can detect specific objects in real-time video streams has the potential to revolutionize various fields and has been the subject of extensive research. Although there have been advances in object detection, there is still a gap in the research for real-time detection of custom objects with high accuracy and speed. This research addresses this gap by training a YOLOv8 detector on a custom dataset of objects and evaluating its performance on real-time video streams which is by far the latest model and thus is faster and more accurate. Our experimental results demonstrate that our custom-trained YOLOv8 detector achieves high accuracy and real-time performance on a custom dataset of objects. The detector achieved an overall mAP50 of 0.864 and a mAP50-95 of 0.758, with individual class results ranging from 0.47 to 0.995. These findings show that custom training data and YOLOv8 are effective in real-time object detection, which has practical applications in various fields. The significance of the results and our contribution lies in demonstrating the effectiveness of custom training data for improving object detection accuracy and speed using YOLO, which has implications for a wide range of real-world applications. © 2023 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

AdaptiveCloset: Reinforcement Learning in Personalized Clothing Recommendations 18

AdaptiveCloset: Reinforcement Learning in Personalized Cloth...

引用

18th IEEE International Conference on Emerging Technologies, ICET 2023

作者： Majeed, Omar Kashif Ul Abideen, Zain Arshad, Usama Ali, Raja Hashim Habib, Abdullah Mustafa, Rafay BCMaterials Basque Center for Materials Applications & Nanostructure Basque Country Leioa Spain Gik Institute of Engg. Sciences & Technology Faculty of Computer Science & Engineering Topi Pakistan University of Europe for Applied Sciences Dept. of Technology & Software Engineering Berlin Germany

ISBN: (纸本)9798350328172

The fashion industry, with its myriad choices, often overwhelms consumers. Addressing this, AdaptiveCloset introduces a groundbreaking approach to tailoring clothing suggestions by harnessing the power of reinforcement learning (RL). Unlike conventional AI methodologies in a fashion that merely suggests based on past preferences, our system dynamically ad-justs using realtime user feedback, ensuring that recommendations remain relevant and personalized. Historically, AI's fusion into the fashion realm has witnessed multiple methodologies but conspicuously lacked the RL perspective. Our research stands at this juncture, aiming to redefine online shopping experiences. Through AdaptiveCloset, we envisage a scenario where online shoppers not only receive personalized recommendations but also feel a sense of involvement, thanks to the system's feedback-oriented adaptability. This responsiveness not only augments user engagement but provides actionable intelligence for businesses, bridging the gap between consumer desires and market offerings. In this study, our focus was to ensure the robustness and adaptability of the RL environment, positioning it as a potent tool for enhancing e-commerce interactions, optimizing sales strategies, and fortifying customer loyalty. © 2023 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Deciphering Faces: Enhancing Emotion Detection with Machine Learning Techniques 18

Deciphering Faces: Enhancing Emotion Detection with Machine ...

引用

18th IEEE International Conference on Emerging Technologies, ICET 2023

作者： Ishaq, Muhammad Haseeb Mustafa, Rafay Arshad, Usama Abideen, Zain Ul Ali, Raja Hashim Habib, Abdullah Gik Institute of Engg. Sciences & Technology Faculty of Computer Science & Engineering Topi Pakistan BCMaterials Basque Center for Materials Applications & Nanostructure Basque Country Leioa Spain University of Europe for Applied Sciences Dept. of Technology & Software Engineering Berlin Germany

ISBN: (纸本)9798350328172

Facial emotion detection holds significant relevance across various domains, from psychology and marketing to education and security. Despite its importance, prevalent techniques often grapple with issues like low precision, susceptibility to lighting changes, obstructions, and distinct facial characteristics. Addressing these challenges, our research embarked on devising a robust and precise facial emotion detector harnessing the potential of machine learning, focusing on convolutional neural networks (CNN). Comprehensive testing revealed that our model surpasses existing state-of-the-art techniques, showcasing superior performance on benchmark datasets. The salience of our research is underscored by its profound implications for myriad real-world applications hinging on accurate facial emotion recognition. We present an enhanced model, distinguished not just by its accuracy but also its robustness, making it apt for diverse scenarios from insightful marketing initiatives and nuanced medical diagnoses to enriched educational experiences. Through this endeavor, we have accentuated the transformative capacity of machine learning in refining and redefining facial emotion detection methodologies. © 2023 IEEE.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

An Enhanced Genetic Algorithm Framework for Efficient Solutions to Capacitated Vehicle Routing Problems 18

An Enhanced Genetic Algorithm Framework for Efficient Soluti...

引用

18th IEEE International Conference on Emerging Technologies, ICET 2023

作者： Arshad, Usama Ellahie, Hamza Hasan Abideen, Zain Ul Ali, Raja Hashim BCMaterials Basque Center for Materials Applications & Nanostructure Basque Country Leioa Spain Gik Institute of Engg. Sciences & Technology Faculty of Computer Science & Engineering Topi Pakistan University of Europe for Applied Sciences Dept. of Technology & Software Engineering Berlin Germany

ISBN: (纸本)9798350328172

Capacitated Vehicle Routing Problems (CVRPs), a widely acknowledged NP-hard issue, pertains to the optimal routing of a limited-capacity vehicle fleet to fulfill customer demand, aiming for the least possible travel distance or cost. Despite the presence of numerous heuristic and exact approaches, the combinatorial characteristic of CVRP renders it challenging, especially for large-scale instances. This research provides an in-depth exploration of utilizing Genetic Algorithms (GAs) to address Capacitated Vehicle Routing Problems (CVRPs), a recognized and intricate optimization issue in the realm of logistics and supply chain management. Our paper concentrates on the innovative usage of GAs, a category of stochastic search methodologies inspired by natural selection and genetics, to grapple with CVRP. We put forth a fresh framework grounded in GA that infuses unique crossover and mutation operations tailor-made for CVRP. Our comprehensive computational trials on benchmark datasets suggest that our GA-centric method is proficient in deriving high-standard solutions within acceptable computational durations, surpassing multiple contemporary techniques concerning solution quality and resilience. Our results also underscore the scalability of our proposed approach, marking it as a viable choice for tackling extensive, real-world CVRPs. This paper enriches the current knowledge bank by demonstrating the prowess of GAs in deciphering complicated combinatorial optimization issues, thus offering a novel viewpoint for future advancements in crafting more robust and efficient CVRP resolutions. © 2023 IEEE.

关键词： Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

DeePMD-kit v3: A Multiple-Backend Framework for Machine Learning Potentials

arXiv

引用

arXiv 2025年

作者： Zeng, Jinzhe Zhang, Duo Peng, Anyang Zhang, Xiangyu He, Sensen Wang, Yan Liu, Xinzijian Bi, Hangrui Li, Yifan Cai, Chun Zhang, Chengqian Du, Yiming Zhu, Jia-Xin Mo, Pinghui Huang, Zhengtao Zeng, Qiyu Shi, Shaochen Qin, Xuejian Yu, Zhaoxi Luo, Chenxing Ding, Ye Liu, Yun-Pei Shi, Ruosong Wang, Zhenyu Bore, Sigbjørn Løland Chang, Junhan Deng, Zhe Ding, Zhaohan Han, Siyuan Jiang, Wanrun Ke, Guolin Liu, Zhaoqing Lu, Denghui Muraoka, Koki Oliaei, Hananeh Singh, Anurag Kumar Que, Haohui Xu, Weihong Xu, Zhangmancang Zhuang, Yong-Bin Dai, Jiayu Giese, Timothy J. Jia, Weile Xu, Ben York, Darrin M. Zhang, Linfeng Wang, Han School of Artificial Intelligence and Data Science Unversity of Science and Technology of China Hefei China AI for Science Institute Beijing100080 China DP Technology Beijing100080 China Academy for Advanced Interdisciplinary Studies Peking University Beijing100871 China State Key Lab of Processors Institute of Computing Technology Chinese Academy of Sciences Beijing100871 China University of Chinese Academy of Sciences Beijing China Baidu Inc. Beijing China Department of Computer Science University of Toronto TorontoON Canada Department of Chemistry Princeton University PrincetonNJ08540 United States University of Chinese Academy of Sciences Beijing100871 China State Key Laboratory of Physical Chemistry of Solid Surfaces iChEM College of Chemistry and Chemical Engineering Xiamen University Xiamen361005 China College of Integrated Circuits Hunan University Changsha410082 China State Key Laboratory of Advanced Technology for Materials Synthesis and Processing Center for Smart Materials and Device Integration School of Material Science and Engineering Wuhan University of Technology Wuhan430070 China College of Science National University of Defense Technology Changsha410073 China Hunan Key Laboratory of Extreme Matter and Applications National University of Defense Technology Changsha410073 China ByteDance Research Beijing100098 China Ningbo Institute of Materials Technology and Engineering Chinese Academy of Sciences Ningbo315201 China College of Materials Science and Opto-Electronic Technology University of Chinese Academy of Sciences Beijing100049 China Key Laboratory of Theoretical and Computational Photochemistry of Ministry of Education College of Chemistry Beijing Normal University Beijing100875 China Department of Geosciences Princeton University PrincetonNJ08544 United States Department of Applied Physics and Applied Mathematics Columbia University New YorkNY10027 United States IKKEM Fujian Xiamen361005 China Graduate

In recent years, machine learning potentials (MLPs) have become indispensable tools in physics, chemistry, and materials science, driving the development of software packages for molecular dynamics (MD) simulations and related applications. These packages, typically built on specific machine learning frameworks such as TensorFlow, PyTorch, or JAX, face integration challenges when advanced applications demand communication across different frameworks. The previous TensorFlow-based implementation of DeePMD-kit exemplified these limitations. In this work, we introduce DeePMD-kit version 3, a significant update featuring a multi-backend framework that supports TensorFlow, PyTorch, JAX, and PaddlePaddle backends, and demonstrate the versatility of this architecture through the integration of other MLPs packages and of Differentiable Molecular Force Field. This architecture allows seamless backend switching with minimal modifications, enabling users and developers to integrate DeePMD-kit with other packages using different machine learning frameworks. This innovation facilitates the development of more complex and interoperable workflows, paving the way for broader applications of MLPs in scientific research. Copyright © 2025, The Authors. All rights reserved.

关键词： Molecular dynamics

来源：评论

学校读者我要写书评

暂无评论

Deep Learning-Based Approaches for Accurate Brain Tumor Detection in MRI Images

Deep Learning-Based Approaches for Accurate Brain Tumor Dete...

引用

2024 International Conference on Decision Aid Sciences and applications, DASA 2024

作者： Noori, Mohammed Ahmed, Mohsin Ibrahim, Amer Saeed, Ali Khishe, Mohammad Mahfuri, Mahmoud Al-ma'Moon University College Baghdad Iraq College of Information Technology Department of Computer Science and Software Engineering United Arab Emirates Northern Technical University Computer Center Ninevah Iraq Yuan Ze University Innovation Center for Artificial Intelligence Applications Taiwan Jadara Research Center Jadara University Jordan International University for Science and Technology Computer and Information Technology Department Syria

ISBN: (纸本)9798350369106

Brain tumor classification is considered one of the major tasks in medical image analysis, where correct and timely diagnosis could be achieved to serve as the key to effective treatment. This research paper proposes a deep learning-based approach for automatic classification of brain tumors from MRI images by fine tuning the ResNet50V2 CNN model. The dataset is made up of four classes, namely: Glioma Tumor, Meningioma Tumor, No Tumor, and Pituitary Tumor, totaling 2,844 images. We have done some strategy data augmentation and class weighting in order to prevent class imbalance, which ensures good performance in all tumor classes. Pre-processing and augmentation of the input images are done via rotation, shifting, zooming, and flipping. It yielded 95.29% on the validation set, with the major highlights of performance metrics for the minority class being class precision equal to 0.97 and class recall equal to 0.92, hence proving the efficiency of performed class balancing. The range of F1-scores, from 0.93 to 0.97, means fantastic predictive capability across all tumor types. Other techniques were also used to make this model optimal, such as early stopping, learning rate reduction, and checkpointing. The high performance of this model shows great promise in helping clinicians toward the right diagnosis of brain tumors in an effective way. Future efforts will involve incorporating larger datasets, exploring advanced augmentation techniques to enhance further model generalizability. Model explainability tools such as Grad-CAM will be used to extract insight into the model's decision-making process. This enhances the clinical interpretability of the results. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Examining OSGi from an ideal enterprise software component model

Examining OSGi from an ideal enterprise software component m...

引用

2010 IEEE International Conference on software engineering and Service Sciences, ICSESS 2010

作者： Zhou, Jingang Zhao, Dazhe Ji, Yong Liu, Jiren National Engineering Research Center for Computer Software Northeastern University 110004 Shenyang China State Key Laboratory of Advanced Software Architecture Neusoft Corporation 110179 Shenyang China School of Management Dalian University of Technology 116023 Dalian China

ISBN: (纸本)9781424460526

Component-based software development is becoming a discipline in software engineering community, but it still immature in enterprise application development. OSGi gives some insights for component-based development (CBD). By proposing an ideal component model for enterprise applications, we analyze the performance of OSGi against such a model. The result shows that OSGi is still lack of some advanced component features for CBD, although it provides a modular solution superior to plain Java. The result also contributes both to the decision making for the software developers who want to select the OSGi component technology and to the improvements to be made to OSGi. © 2010 IEEE.

关键词： Enterprise software

来源：评论

学校读者我要写书评

暂无评论

Slide-based Graph Collaborative Training for Histopathology Whole Slide Image Analysis

引用

IEEE Transactions on Medical Imaging 2025年

作者： Shi, Jun Shu, Tong Jiang, Zhiguo Wang, Wei Wu, Haibo Zheng, Yushan Hefei University of Technology School of Software Hefei230601 China Hefei University of Technology School of Computer Science and Information Engineering Hefei230601 China School of Astronautics Beihang University Image Processing Center Beijing102206 China Beihang University Tianmushan Laboratory Zhejiang Hangzhou311115 China University of Science and Technology of China Division of Life Sciences and Medicine Department of Pathology the First Affiliated Hospital of USTC Hefei230036 China University of Science and Technology of China Division of Life Sciences and Medicine Intelligent Pathology Institute Hefei230036 China Beihang University School of Engineering Medicine Beijing Advanced Innovation Center on Biomedical Engineering Beijing100191 China

The development of computational pathology lies in the consensus that pathological characteristics of tumors are significant guidance for cancer diagnostics. Most existing research focuses on the inner-contextual information within each WSI yet ignores the possible inter-correlations between slides. As the development of tumors is a continuous process involving a series of histological, morphological, and genetic changes that accumulate over time, the similarities and differences between WSIs across various stages, grades, locations and patients should potentially contribute to the representation of WSIs and deserve to be taken into account in WSI modeling. To verify the advancement of introducing the slide inter-correlations into the representation learning of WSIs, we proposed a generic WSI analysis pipeline SlideGCD that can be adapted to any existing Multiple Instance Learning (MIL) frameworks and improve their performance. With the new paradigm, the prior knowledge of cancer development can participate in the end-to-end workflow, which concurrently initializes and refines the slide representation, as a guide for message passing in the slide-based graph. Extensive comparisons and experiments are conducted to validate the effectiveness and robustness of the proposed pipeline across 4 different tasks, including cancer subtyping, cancer staging, survival prediction, and genemutation prediction, with 8 representative SOTA WSI analysis frameworks as backbones. © 1982-2012 IEEE.

关键词： Image correlation

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：