检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Wei, Hao Zhou, Yanhui Jia, Yiwen Ge, Chenyang Anwar, Saeed Mian, Ajmal National Key Laboratory of Human-Machine Hybrid Augmented Intelligence Institute of Artificial Intelligence and Robotics Xi’an Jiaotong University Xi’an710049 China School of Information and Telecommunication Xi’an Jiaotong University Xi’an710049 China Department of Computer Science and Software Engineering The University of Western Australia Crawley PerthWA6009 Australia

Perceptual image compression has shown strong potential for producing visually appealing results at low bitrates, surpassing classical standards and pixel-wise distortion-oriented neural methods. However, existing methods typically improve compression performance by incorporating explicit semantic priors, such as segmentation maps and textual features, into the encoder or decoder, which increases model complexity by adding parameters and floating-point operations. This limits the model’s practicality, as image compression often occurs on resource-limited mobile devices. To alleviate this problem, we propose a lightweight perceptual Image Compression method using Implicit Semantic Priors (ICISP). We first develop an enhanced visual state space block that exploits local and global spatial dependencies to reduce redundancy. Since different frequency information contributes unequally to compression, we develop a frequency decomposition modulation block to adaptively preserve or reduce the low-frequency and high-frequency information. We establish the above blocks as the main modules of the encoder-decoder, and to further improve the perceptual quality of the reconstructed images, we develop a semantic-informed discriminator that uses implicit semantic priors from a pretrained DINOv2 encoder. Experiments on popular benchmarks show that our method achieves competitive compression performance and has significantly fewer network parameters and floating point operations than the existing state-of-the-art. We will release the code and trained models. Copyright © 2025, The Authors. All rights reserved.

关键词： Decoding

来源：评论

学校读者我要写书评

暂无评论

Energy-Efficient Resource Allocation in Wireless Networks Using Federated-Double-Deep Q-Network

Energy-Efficient Resource Allocation in Wireless Networks Us...

引用

Integrated Circuits and Communication Systems (ICICACS), IEEE International Conference on

作者： P. Anuradha Muhamed Fallahhusein Sakthivel Subramaniam T. Aditya Sai Srinivas A. Buckshumiyan Department of Electronics and Communication Engineering Chaitanya Bharathi Institute of Technology Hyderabad India Department of Computers Techniques Engineering College of Technical Engineering The Islamic University Najaf Iraq Department of Computer Science and Engineering Sona College of Technology Salem India Department of Artificial Intelligence and Machine Learning Jayaprakash Narayan College of Engineering Mahbubnagar India Department of Mechanical Engineering New Prince Shri Bhavani College of Engineering and Technology Chennai India

ISBN: (数字)9798331508456

ISBN: (纸本)9798331508463

Now a days, the resource allocation manages the resource networks such as bandwidth, power usage for ensuring Quality of Service (QoS) without overloading. However, the optimization fails to adapt traffic loads due to its dependance on static resource allocation that cannot adjust to user demands and power control. To overcome this limitation in this research resource allocation method Federated-Double-Deep Q-Network (F-DDQN) is proposed for maximizing throughput and achieve minimum QoS requirements for all active Device-to-Device communication(D2D) pairs and Cellular Users (CUE). whereas Federated learning enhances adaptability to dynamically communication environments. Initially, the data is collected from D2D-enabled heterogeneous cellular networks (HetNet) dataset and 3-tier HetNet system model are designed. Further, for resource management the F-DDQN enhances the energy-efficient by decentralizing action states and minimizes communication overhead and power consumption by sharing model updates instead of raw data, and DDQN effectively improves the energy efficiency in resource allocation by estimating Q-values and optimizes resource allocation through selecting the most energy-efficient transmitting power and channels using target estimation. From the results the proposed model achieved better results to evaluate its effectiveness in increasing system throughput with 58.90% compared with both CUE and D2D Users Equipment (DUE) with existing method a Deep Q-network-based Reinforcement learning algorithm (DRL).

关键词： Federated learning Wireless networks Estimation Quality of service Throughput Dynamic scheduling Energy efficiency Data models Device-to-device communication Resource management

来源：评论

学校读者我要写书评

暂无评论

Design Thinking in Electrical engineering: A Pathway to Human-Centered Innovations

Design Thinking in Electrical Engineering: A Pathway to Huma...

引用

Electronics and Renewable Systems (ICEARS), International Conference on

作者： D. Kavitha M. Siva Ramkumar Arun M Rahmath Ulla Baig M. Sivaramkrishnan Dept of CSE (IoT) SNS College of Engineering India Dept of ECE SNS College of Technology India Department of Mechanical Engineering Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences (SIMATS) Saveetha University Chennai India Department of Artificial Intelligence and Machine Learning Cambridge Institute of Technology K R Puram Bangalore Dept of EEE Karpagam COllege of Engineering

ISBN: (数字)9798331509675

ISBN: (纸本)9798331509682

Design thinking is a useful technique in many technical domains since it is a novel methodology that emphasizes comprehending and meeting customer demands. Using design thinking in electrical engineering provides a fresh approach to creating user-focused solutions in a field where technological complexity is constantly changing. In order to foster creativity and useful problem-solving, this study explores how applying design thinking concepts might transform electrical engineering's conventional procedures. Engineers may produce more efficient and user-friendly designs by utilizing crucial phases like empathy, problem identification, ideation, prototyping, and iterative testing. In order to demonstrate how this strategy results in more flexible and sustainable solutions, the discussion cites particular instances, such as developments in consumer electronics, electric power systems, and renewable energy. The study also emphasizes how teamwork and ongoing iteration foster creativity while maintaining designs' alignment with user requirements and societal impact.

关键词： Electrical engineering Renewable energy sources Technological innovation Smart cities Transforms User experience Teamwork Iterative methods Creativity Testing

来源：评论

学校读者我要写书评

暂无评论

Linking Unknown Characters Via Oracle Bone Inscriptions Retrieval

SSRN

引用

SSRN 2023年

作者： Chen, Xu Li, Bang Liu, Yongge Han, Yahong Jiang, Runhua Li, Deng The College of Intelligence and Computing Tianjin University Tianjin300072 China Tianjin Key Lab of Machine Learning Tianjin300072 China The School of Computer and Information Engineering Anyang Normal University Henan Anyang455000 China

Searching for related information among numerous raw materials (rubbing images) is a crucial step in deciphering unknown Oracle Bone Inscription (OBI) characters. Existing OBI information processing systems rely on databases compiled by experts, which is a time-consuming and labor-intensive procedure. More importantly, they cannot query relevant information about unknown characters inherently. In this paper, we propose a deep-learning-based OBI retrieval framework, dubbed LUC, to enable searching for arbitrary oracle characters, regardless of whether they have been deciphered. We propose an additional domain-aware embedding module to address the significant domain gap between clean glyphs and rubbing images. Besides, we introduce the principle of maximal coding rate reduction into metric learning. A tailored OBI retrieval benchmark is presented and the results show that LUC achieves superior performance compared to other retrieval methods. Furthermore, experiments on three public person ReID benchmarks also verify the effectiveness and generalization of our method. © 2023, The Authors. All rights reserved.

关键词： Image retrieval

来源：评论

学校读者我要写书评

暂无评论

Privacy-Preserving Data Mining Methods for Sensitive Information

Privacy-Preserving Data Mining Methods for Sensitive Informa...

引用

Technological Advancements in Computational Sciences (ICTACS), International Conference on

作者： Sukhvinder Singh L.H. Dari Jasim Vijay Dhote D. Suseela R. Venkatasubramanian Symbiosis Law School Nagpur Symbiosis International (Deemed University) Pune India College of Technical Engineering The Islamic University Najaf Iraq Department of Computer Science & Engineering IES College of Technology Bhopal India Department of Artificial Intelligence and Machine Learning Bannari Amman Institute of Technology Sathyamangalam Department of Computer Science & Engineering New Prince Shri Bhavani College of Engineering and Technology India

This study explores innovative Privacy-Preserving Data Mining (PPDM) methods for handling sensitive information, addressing the critical balance between data utility and privacy. We identify prevalent issues arising from traditional data mining techniques and evaluate advanced PPDM approaches like differential privacy, homomorphic encryption, and federated learning using real-world datasets. Through a comprehensive analysis, our results indicate that hybrid models, which combine multiple PPDM techniques, are particularly effective in enhancing privacy without compromising data utility. The findings underscore the need for adaptable and robust PPDM strategies in the era of Big Data, contributing valuable insights to both the academic and practical domains. Our research paves the path for future studies and developments in creating a harmonious ecosystem where data mining efficiency and privacy coexist, promoting an insightful yet private data-centric world.

关键词：

来源：评论

学校读者我要写书评

暂无评论

E-Commerce Product Review Classification based on Supervised machine learning Techniques

E-Commerce Product Review Classification based on Supervised...

引用

International Conference on Advanced Computing and Communication Systems (ICACCS)

作者： Deepak Dharrao Sarika Deokate Anupkumar M Bongale Siddhaling Urolagin Department of Computer Science and Engineering Symbiosis Institute of Technology Symbiosis International (Deemed University) Pune India Department of Computer Science and Engineering Pimpari Chinchwad College of Engineering Pune India Department of Artificial Intelligence and Machine Learning Symbiosis Institute of Technology Symbiosis International (Deemed University) Pune India Department of Computer Science and Engineering Birla Institute of Technology & Science Dubai

To become a successful organization it is the most important aspect to consider is the customer satisfaction. A business can grow only when the customers enjoy their product/service and how much they recommend it. For the same, companies collect reviews and ratings from its customers to understand their perspective better. Amazon is also such a giant, which values its customers’ opinions. It is very difficult or nearly impossible to have a detailed manual scan of each and every review for each and every product available on the website. This is where NLP (Natural language processing) comes into picture. It processes human language from text to arrays of numbers in order to find sense in the text so that the machine learning algorithms can process them further and classify the texts. This project focuses on processing the amazon reviews of various categories and products and then classify them into 5 ratings, from 1 to 5, worst to best respectively. The supervised machine learning algorithms used to process these reviews are Support Vector machines (SVM), Naïve Bayes classifier and Decision Tree classifier. The analysis of comparison of these algorithms has also been discussed using accuracy score bar charts.

关键词： Sentiment analysis machine learning algorithms Support vector machine classification machine learning Manuals Tag clouds Decision trees

来源：评论

学校读者我要写书评

暂无评论

An Efficient Biomedical Solicitation in Liver Cancer Classification by Deep learning Approach

An Efficient Biomedical Solicitation in Liver Cancer Classif...

引用

Integrated Circuits and Communication Systems (ICICACS), IEEE International Conference on

作者： N Vasundhara Archana S Nandan S V Hemanth Sivudu Macherla Madhura G K Department of CSE CMR Technical Campus Hyderabad India Department of Computer Science & Engineering GITAM School of Technology GITAM University India Department of Computer Science and Engineering Hyderabad Institute of Technology and Management Telangana India PACE Institute of Technology & Science Andhra Pradesh India Department of Artificial Intelligence and Machine Learning Nitte Meenakshi Institute of Technology Bengaluru India

The liver, a significant physical organ, is located in the upper right and left abdominal cavities just below the diaphragm. It produces several different chemicals that the body needs to function properly. Researchers can benefit from converting images into data to more easily exchange and generate precise results. Since the process of conversion relies on technology and an algorithm, it eliminates the possibility of human error. Liver cancer has the highest fatality rate of any cancer since its symptoms don't present until late in the disease's progression, making early detection difficult. Skimming, sifting, segmenting, feature abstraction, and presentation via ANN (artificial Neural Network) are the primary topics of the first stage of the debate. Real-time data sets use Feed-Forward Neural Network (FFNN) for identifying liver cancer and classifying its severity. Filtering has two primary applications: noise suppression and edge rounding. Then, segmentation is employed to isolate the relevant area, allowing for more compact data storage. The Gray Level and Co-occurrence Matrix (GLCM) is used to extract features, and the resulting matrix can have many different forms. This criterion helps classify tumors as benign or malignant. Metrics such as accuracy, sensitivity, positive and negative predictive values, and precision are used to assess the rate. The experimental method for identifying liver tumors uses CT liver pictures to achieve an average accuracy of 99.45% for malignant images.

关键词： Liver cancer Sensitivity Computed tomography Liver Biological systems artificial neural networks Watersheds

来源：评论

学校读者我要写书评

暂无评论

Prognostic Prediction for Non-small-Cell Lung Cancer Based on Deep Neural Network and Multimodal Data 17th

Prognostic Prediction for Non-small-Cell Lung Cancer Based o...

引用

17th International Conference on Intelligent Computing, ICIC 2021

作者： Zhang, Zhong-Si Xu, Fei Jiang, Han-Jing Chen, Zhan-Heng Key Laboratory of Image Information Processing and Intelligent Control of Education Ministry of China Institute of Artificial Intelligence School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan430074 China Institute of Machine Learning and Systems Biology School of Electronics and Information Engineering Tongji University No. 4800 Caoan Road Shanghai201804 China

ISBN: (纸本)9783030845315

Non-small-cell lung cancer (NSCLC) is the most common lung cancer with poor prognosis. Prognostic prediction is significant in improving the prognosis of NSCLC patients. Clinical information and multi-omics data including gene expression, miRNA, copy number variations, and DNA methylation are closely related to NSCLC prognosis. In this study, we propose a deep neural network to conduct prognostic prediction for NSCLC patients based on all five types of data. Given the high dimensional features of the omics data, past works reduce the feature dimension by regression analysis or correlation sorting algorithms. A shortcoming of these methods is that only a small number of features are considered. To overcome it, we propose a convolution neural network-based feature transformation method, which considers all features and extracts the abstract representations of the omics data. Based on the representations, we predict the five-year survival status of NSCLC patients. The results show that our method achieves a more precise prediction than previous work, with the areas under the curve improved from 0.8163 to 0.8508 and the accuracy improved from 0.7544 to 0.8096. © 2021, Springer Nature Switzerland AG.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Digital Evidence Management: Modern Techniques in Digital Forensics

Digital Evidence Management: Modern Techniques in Digital Fo...

引用

Electronics and Renewable Systems (ICEARS), International Conference on

作者： N. Venkataramanan Jayalakshmi V S. Anitha R. Ganesan V. Jenifer Department of Computer Applications DRBCCC Hindu College Chennai Department of Computer Science and Applications SRM Institute of Science and Technology (FSH) Ramapuram Department of Artificial Intelligence and Machine Learning St. Joseph’s College of Engineering Anna University Chennai Department of Computer Science and Engineering Vel Tech Rangarajan Dr.Sagunthala R&D Institute of Science and Technology Chennai Department of Computer Science and Applications SRM Institute of Science and Technology (FSH)

ISBN: (数字)9798331509675

ISBN: (纸本)9798331509682

This research study reviews the significance of digital forensic analysis in modern criminal investigations, highlighting the way it complements conventional methods of inquiry and the difficulties it encounters in the information era. To solve crimes effectively, investigators must combine digital knowledge with traditional forensic techniques. This will enable them to use evidence from many different types of digital and physical sources. The specific intricacies of digital crime scenes need specialized methods for the proper storage and analysis of digital evidence to identify criminal intent and patterns of conduct. To propose Modern Techniques for Digital Evidence Management [MT-DEM] based Digital Forensics Service. Multidisciplinary teams and cutting-edge technology are becoming necessities in digital forensic investigation as the field adjusts to new threats like data encryption and massive amounts of data. The admissibility of evidence in the legal world depends on digital forensic procedures being transparent and adhering to set criteria. Digital forensic analysis has become an indispensable tool for contemporary investigators with the rise of digital evidence.

关键词： Renewable energy sources Analytical models Systematics Law enforcement Reviews Digital forensics Companies Encryption Internet of Things Best practices

来源：评论

学校读者我要写书评

暂无评论

GDM-BC: Non-invasive body composition dataset for intelligent prediction of Gestational Diabetes Mellitus

引用

computers in Biology and Medicine 2025年 192卷 110176-110176页

作者： Zheng, Chen Qing, Tong Li, Mao Liao, Shujuan Luo, Biru Tang, Chenwei Lv, Jiancheng College of Computer Science Sichuan University Chengdu610065 China Engineering Research Center of Machine Learning and Industry Intelligence Ministry of Education Chengdu610065 China Department of Nursing West China Second University Hospital Sichuan University/West China School of Nursing Chengdu610041 China Ministry of Education Chengdu610041 China

Gestational Diabetes Mellitus (GDM) refers to any degree of impaired glucose tolerance with onset or first recognition during pregnancy. As a high-prevalence disease, GDM damages the health of both pregnant women and fetuses in the short and long term. Accurate and cost-effective recognition of GDM is quite crucial to reduce the risk and economic pressure of this disease. However, existing datasets for the prediction of GDM primarily focus on clinical and biochemical parameters, including a mass of invasive indexes. These variables are hard to obtain and do not always perform well in the prediction of GDM. In this paper, we introduce a large-scale non-invasive body composition dataset, called GDM-BC, for intelligent risk prediction of GDM. Specifically, it contains a cohort of 39,438 pregnant women, of whom 7777 (19.7%) were subsequently diagnosed with GDM. Besides, our dataset includes a large number of body composition indexes that can be acquired non-invasively. In addition, we perform several traditional machine learning and deep learning methods on the GDM-BC dataset, among which the Residual Attention Fully Connected Network (RAFNet) performs the best, achieving an AUC (area under the ROC curve) of 0.920. The results show that our dataset is marvelous and creates a new perspective on the prediction of GDM. Our models may offer an opportunity to establish a cost-effective screening approach for identifying low-risk pregnant women based on body composition data. We believe that our proposed GDM-BC dataset will advance future research on risk prediction for GDM, as well as provide new insights for intelligent prediction of other high-incidence pregnancy-related diseases such as gestational hypertension. © 2025 Elsevier Ltd

关键词： Prediction 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：