检索结果-内蒙古大学图书馆

Multimodal search in fashion store using CLIP

AIP Conference Proceedings 2024年第1期3134卷

作者： Isaac Elizabeth Anu Mathew Joby Roy Meckamalil Rotney Sagar V. K. Sruthi N. Kartha Sreekanth P. M. Sreelakshmi Assistant Professor Department of Computer Science & Engineering Mar Athanasius College of Engineering Kothamangalam Kerala India Department of Computer Science & Engineering Mar Athanasius College of Engineering Kothamangalam Kerala India

Contemporary computer vision systems are typically trained to recognize a limited set of predetermined object categories. This restricted type of supervision limits their applicability and generalization since additional labeled data is needed to indicate any other visual concept. A potential approach that leverages a broader source of supervision involves learning directly from raw text about images. The proposed solution involves implementing a multimodal search in a fashion store using the Contrastive Language Image Pretraining (CLIP) model, which enables vector-based search rather than syntax-based search. CLIP improves product visibility and assists users in obtaining better search results that match their preferences. It is demonstrated that a simple pre-training task of predicting the corresponding image for a given caption is an effective and scalable technique for learning state-of-the-art image representations from scratch using a dataset of (image, text) pairs collected from the internet. Natural language is employed during pre-training to refer to learned visual concepts or describe new ones, enabling zero-shot transfer of the model to downstream tasks. Integration with clustering techniques on the vector space further enhances CLIP’s capabilities, facilitating efficient and accurate retrieval of items. By grouping similar fashion items based on visual similarity and semantic relationships, the system enables efficient and accurate retrieval of relevant items. The model can generalize effectively to most tasks without requiring any training on a specific dataset and is frequently on par with a fully supervised baseline.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Deep Learning and its Applications: A Novel Approach to Machine Learning with Encoding Algorithms

Deep Learning and its Applications: A Novel Approach to Mach...

引用

Intelligent & Innovative Practices in engineering & Management (IIPEM), International Conference on

作者： Kamalesh Thoppaen Suresh Babu Bhagyashree Ashok Tingare Vijayshri Khedkar Tarun Dhar Diwan P. William Abhishek Badholia Globallogic Inc Santa Clara CA USA Department of Artificial Intelligence and Data Science D Y Patil College of Engineering Akurdi Pune Department of Computer Science & Engineering Symbiosis Institute of Technology Symbiosis International (Deemed University) Pune India Assistant Professor and Controller of Examination (COE) Atal Bihari Vajpayee University Bilaspur India Deaprtment of Information Technology Sanjivani College of Engineering Kopargaon Amity University Dubai UAE Department of Data Science Shri Shankaracharya Institute of Professional Management and Technology Raipur

ISBN: (数字)9798350390049

ISBN: (纸本)9798350390056

Deep learning has revolutionized data processing by recognizing complicated patterns using multi-layered neural networks. Encoding complex data like hyperspectral photos may help deep learning models perform better. Encoding shapes raw data for compression, dimensionality reduction, and feature extraction. These methods matter. This research found that higher-dimensional encodings can improve deep learning model accuracy and convergence. Complex encoding algorithms like FE-16D and BLOSUM-20D outperform simpler ones in validation accuracy and AUC. To prove this, extensive testing and comparison were done. The study emphasizes the necessity to combine deep learning frameworks with improved encoding techniques to improve machine learning and large-scale data processing.

关键词： Deep learning Dimensionality reduction Machine learning algorithms Computational modeling Neural networks Feature extraction Data processing Encoding Data models Hyperspectral imaging

来源：评论

学校读者我要写书评

暂无评论

Developing a computer-Aided Diagnostic System for Breast Cancer Ultrasound Imaging

Developing a Computer-Aided Diagnostic System for Breast Can...

引用

Data science and engineering in Healthcare, Medicine and Biology, IEEE EMBS Special Topic Conference on

作者： Radwa Taha Shereen Afifi Mohamed A. Abd-ElGhany Mohammed A.-M Salem Master Candidate in the Faculty of Media Engineering and Technology German University in Cairo New Cairo Egypt Assistant Professor in Computer Science Engineering Faculty of Media Engineering and Technology German University in Cairo New Cairo Egypt Professor in Electrical Engineering Faculty of Information Engineering and Technology German University in Cairo New Cairo Egypt Integrated Electronics Systems Lab TU Darmstadt Germany Digital Media Engineering Faculty of Media Engineering and Technology German University in Cairo New Cairo Egypt

In this research we aim to develop a computer-aided diagnostic system (CAD) for breast ultrasound imaging to enhance the early detection of breast cancer. CAD tools aid physicians in the diagnostic process leading to early detection and treatment of breast cancer. The proposed CAD tool comprises three stages, the first is enhancement of breast ultrasound imaging by applying custom segmentation deep learning architectures [U-Net, SK-U-Net, RDA-U-Net] to enhance the isolation of masses in Ultrasound images, then generation of new ultrasound images by employing Generative Adversarial Networks [DAGAN, DCGAN] to overcome data limitation in available data-sets, and the last stage is the detection and classification of breast cancer masses by a CNN model. The enhancement stage shows promising results with accuracy of 98 % overall accuracy, 92 % overall precision, and 85 % overall *** Relevance: Breast cancer is one of the deadly diseases in the world. The earlier the detection the higher the chances of recovery. Thus a CAD tool that facilitates the screening process of ultrasound images will aid physicians in the early detection and diagnosis of the disease.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Automated data preprocessing and training interface for machine learning applications

引用

AIP Conference Proceedings 2024年第1期2942卷

作者： Anagha Challa Akhila Vutukuri Maithreyi Kenguva Sateesh Kumar Kanagala Department of Electronics and Computer Engineering Sreenidhi Institute of Science and Technology Hyderabad India Assistant Professor Department of Electronics and Computer Engineering Sreenidhi Institute of Science and Technology Hyderabad India

Machine learning is a promising tool for analyzing and predicting data. However, it is very time-consuming and challenging to manually process the data. There are also many missing values in the data that can affect the accuracy of the models. We propose an automated approach to pre-process the data. It efficiently performs the various steps involved in the data cleaning process by converting the categorical values to a label encoded value, sampling the data, and replacing the missing values with the appropriate central tendency. This project takes the raw data as input and provides a clean error-free dataset that is suitable for training machine learning models. It also aims to use a web application to automate data pre-processing and training of classification and regression models. It uses multiple algorithms and normalization strategies to get the optimum model based on the desired metric. It allows users to download the clean dataset as well as the trained models after the models have been trained. Multiple datasets, including binary classification and regression, are used to test this model.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Game theory-based optimization for efficient IoT task offloading in 6G network base stations

引用

Measurement: Sensors 2024年 33卷

作者： Keshta, Ismail Soni, Mukesh Deb, Nabamita singh, Shweta Saravanan, K. Raza Khan, Dr Ihtiram Computer Science and Information Systems Department College of Applied Sciences AlMaarefa University Riyadh Saudi Arabia Dr. D. Y. Patil Vidyapeeth Pune Dr. D. Y. Patil School of Science & Technology Tathawade Pune India Assistant Professor Department of Information Technology Gauhati University India Department of Information Technology R.M.D. Engineering College Chennai India Computer Science Department Jamia Hamdard Delhi India Electronics and communication department IES college of technology Bhopal India

This study aims to optimise computing for intricate jobs within the overlapping coverage of 6G network base stations. A multi-access edge computing network model is created by solving the issues of task offloading. This model involves many base stations and IoT devices. It takes into account factors such as task delay, energy consumption, societal impacts, and economic incentives. The objective of joint optimization is to maximize the revenues of base stations and the utilities of IoT devices by considering base station price, IoT device base station selection, and job offloading mechanisms. The IoT device base station selection is addressed using a many-to-one matching game model, while pricing and task offloading interactions are managed by a two-stage approach based on Stackelberg game theory. The suggested Optimal Pricing and Best Response Algorithm (OBGT), which is based on Game Theory, successfully reaches equilibrium solutions. It demonstrates rapid convergence in simulations and improves both base station profits and IoT device utility. This study integrates data-driven mobile computing systems assurance for futuristic smart vertical networks, enhancing the efficiency of task offloading and optimization. © 2024 The Authors

关键词： Game theory

来源：评论

学校读者我要写书评

暂无评论

Strength analysis of geopolymer concrete based on fly ash and P-sand

引用

Materials Today: Proceedings 2021年 47卷 5489-5492页

作者： R. Rajendrana D. Sarath Kumar K. Megan P. Muneswaren D. Senthilvel B. Shri N. Lingeshwaran Assistant Professor Department of Civil Engineering K Ramakrishnan College of Technology Trichy 621112 India Research Scholar SRM Institute of Science and Technology Kattankukathur TN India UG Student Department of Civil Engineering K Ramakrishnan College of Technology Trichy 621112 India Assistant Professor Department of Civil Engineering Koneru Lakshmaiah Education Foundation (Deemed to be University) Guntur Andhra Pradesh 522502 India

The genuine emergency of the world is confining today for trending of natural causes the ecological contamination. In the development business principally the creation of OPC will causes the discharge of toxins brings about ecological contamination. We can cut down the contamination impact on climate, by broadening the utilization of modern outgrowth in our development industry. Geopolymer concrete is the current study, to create the geo-polymer concrete. The Portland concrete is completely replace with fly debris and the fine total is replaced with p-sand and basic fluids can be utilized for the limiting of materials. The antacid fluids utilized in this assessment for the polymerization are the arrangements of Sodium hydroxide (NaOH) and sodium silicate (Na2Sio3). Various molarities of sodium hydroxide arrangement for example 8 M, 10 M and 12 M are taken to get ready distinctive mix. Also, the compressive strength is determined in the structural lab. The solid shape examples are taken of size 150 mm × 150 mm × 150 mm and the 1 m 3 3D square. The geopolymer concrete specimens are tried for their compressive strength at the period of 21 days, mixes of different sodium hydroxide molarities i.e.8 M, 10 M and 12 M are readied and they are cured by direct daylight and their strengths are determined for 21 days. The outcome shows that the strength of geopolymer concrete rises with the expansion of the molar concentration of sodium hydroxide.

关键词： Geopolymer concrete Sodium hydroxide Sodium silicate Compression test Split tensile and flexural test

来源：评论

学校读者我要写书评

暂无评论

Comparative machine learning analysis of microalgae and biogas blended fuel in Micro-gas turbine using long short-term memory and convolutional neural network approaches

引用

Journal of the Taiwan Institute of Chemical Engineers 2024年

作者： T R, Praveenkumar S, Prakash Alharbi, Sulaiman Ali Almoallim, Hesham S. M, Shobana Department of Civil Engineering Graphic Era Deemed to be University Uttarakhand Dehradun248002 India Professor & Principal Mahalakshmi Tech Campus Chromepet Chennai India Department of Botany and Microbiology College of Science King Saud University PO Box -2455 Riyadh 11451 Saudi Arabia Department of Oral and Maxillofacial Surgery College of Dentistry King Saud University PO Box -60169 Riyadh 11545 Saudi Arabia Department of Computer Science TMG College of Arts & Science Chennai India

Background: The increasing demand for energy-efficient and environmentally friendly energy systems has necessitated the exploration of advanced machine learning techniques in predicting key engine parameters. The study addresses the challenge of optimizing turbine inlet temperature and EGT using machine learning technique that integrates various fuel types, including microalgae. In the current study, the machine learning approach has been used to predict the effects of the biofuel and biogas in the micro gas turbine engine. Utilization of the biogas in the micro gas turbine (MGT) has been a major barrier due to the complexity in the combustor design and its reaction mechanism. Direct injection of the biogas into the MGT needs a major design modification to the fuel injector design. This study takes an alternate approach to inject the biogas into the MGT via inlet manifold of the MGT. Methods: The experimental test rig has been modified slightly to handle the biogas injection. Five different types of fuel had been tested, with various fraction of microalgae-based biofuel with biogas. Biogas acts as the energy carrier to stimulate the combustion inside the combustion chamber and reduces the emission. Blends were tested with different turbines speeds and the performance of the MGT along with emissions was measured. The obtained data was then processed using regression models implemented through LSTM and CNN architectures. Significant Findings: From the findings, use of the microalgae blends in the MGT is more appealing when the biogas has been injected. There is a major shift in both performance and emissions were noted. There is a slight decrease in the static thrust observed irrespective of the microalgae concentration in the fuel due to the lower energy content. As the energy content more fuels were supplied in to the combustion process to avoid flameout in the MGT. Hence all blends reported the highest thrust specific fuel consumption than Jet-A. This pattern has b

关键词： Gas turbines

来源：评论

学校读者我要写书评

暂无评论

AI-Based Cybersecurity Frameworks for 7G-Enabled Virtual Therapy Platforms

Cyber Security and Applications

引用

Cyber Security and Applications 2025年

作者： A. Jayanthiladevi Jayapandian Natarajan KP Arjun L. Godlin Atlas M. Arvindhan Daniel Arockiam Professor Department of Computer Engineering Marwadi University Rajkot Gujarat India Associate Professor Department of Computer Science and Engineering School of Engineering and Technology Christ University Kengeri Campus Bangalore Karnataka India Department of Computer Science and Engineering GITAM University Bengaluru Campus Karnataka India Associate Professor Department of Computer Science and Engineering Bharath Institute of Higher Education and Research Chennai Tamil Nadu Assistant Professor Amity School of Engineering & Technology Computer Science and Engineering Gwalior Madhya Pradesh India Associate Professor CSE Amity University Madhya Pradesh Gwalior 474005 India

The modern technology network plays an important role. Artificial intelligence to provide a higher solution for network security. As 7G networks emerge, virtual therapy platforms have begun to take center stage, offering fast, real-time interaction and high-definition immersive experiences. While the convergence of security in health care and mobility is in many ways even more integrated into our lives, it is fraught with cybersecurity risks. This article proposes an AI-based cybersecurity framework for a 7 G-based virtual therapy platform. The framework employs advanced machine learning (ML) algorithms, predictive analytics, and adaptive threat intelligence to protect the confidentiality, integrity, and availability of data. The article covers the core building blocks like anomaly detection, secure 7 G-based virtual therapy platform authentication protocols, and privacy-preserving techniques. AI-GN gives an overview of attack mitigation using AI with a comparative analysis of existing 5G/6G frameworks and solutions and their unique aspects needed for 7G environments. This article mainly focuses on threat detection using a deep learning method, also it aims to provide a continuous authentication process. The proposed model response time is decreased by nearly 30 percent, at the same time detection rate is increased to 98.5 percent in a 7G network.

关键词： AI-based cyber security framework Adaptive threat intelligence Anomaly detection Predictive analytics Threat detection

来源：评论

学校读者我要写书评

暂无评论

Survey on various DDoS attacks on firewall and study on emerging DoF attacks

引用

AIP Conference Proceedings 2023年第1期2773卷

作者： P. A. Jishiya Ajay James K. P. Swaraj M-Tech Student Computer Science and Engineering Govt. Engineering College Thrissur India Assistant Professor Computer Science and Engineering Govt. Engineering College Thrissur India Associate Professor Computer Science and Engineering Govt. Engineering College Thrissur India

A firewall is a network device that act as a security guard between a trusted network and an untrusted network. Firewall is the first line of defence against malicious traffic. Nowadays, attacking an organisation network for evil intentions has become a common thing. Distributed Denial of Service (DDoS) attack floods network traffic with illegitimate traffic and thereby preventing accessibility to legitimate users. However new type of attacks are being emerged targeting the network devices and their vulnerabilities. Attacker uses specially crafted packets to overload firewall with comparatively lesser effort, known as Denial of Firewalling (DoF) attack. This paper discusses about the attacks targeting the firewall and also about the newly emerged DoF attack. Moreover, a detection model for DDoS attack is also analysed.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Comparative analysis of reduced feature set based ML models for IDS

引用

AIP Conference Proceedings 2023年第1期2773卷

作者： P. A. Jishiya K. P. Swaraj Ajay James M-Tech Student Computer Science and Engineering Govt. Engineering College Thrissur India Associate Professor Computer Science and Engineering Govt. Engineering College Thrissur India Assistant Professor Computer Science and Engineering Govt. Engineering College Thrissur India

Cyber-attacks targeting organisations and individuals are getting worse day by day. With the technological improve-ments, the principles used to launch attacks are also varied. Intrusion Detection System (IDS) is a network solution to detect attacks and anomalies in the network. Leveraging Machine Learning for developing an IDS can make it more powerful. This work aims at evaluating the performance of machine learning models for Intrusion Detection System built based on CART Algorithm, ID3Decision tree algorithm, Random forest algorithm with Convolutional Neural Network. The model is built using KDD Cup99 Dataset. In this work we perform Linear Correlation and Mutual Information feature selection algorithms for ML algorithms and for Convolutional Neural Network algorithm being a deep learning technique feature selection happens automatically. Analysing various parameters including F1-score, ID3 has shown better results.

关键词：

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：