检索结果-内蒙古大学图书馆

Diabetic retinopathy classifier with convolution neural network

Materials Today: Proceedings 2023年 72卷 1765-1773页

作者： Ingle, Vaishali Ambad, Prashant Department of Computer Science & Information Technology Dr. Babasaheb Ambedkar Marathwada University Maharashtra Aurangabad 431 004 India Department of Mechanical Engineering Maharshtra Institute of Technology Maharashtra Aurangabad 431 010 India

Problems in vision that are originated by diabetes are known as Diabetic Retinopathy (DR).The requirement for a inclusive and automatic method of DR test is raised by doctors. Various methods such as classification, recognition of images are in the past used for DR detection. Automated detection techniques used for early detection can save eyes of diabetic patients. Two deep learning models using Convolution Neural Network are designed and trained on images. The models are tested on dataset collected from Kaggle and APTOS repository, comprising 3662 training and 1928 testing images. The accuracy of designed models is calculated it is approximately 93–95% for given dataset. The model results are cross verified by expert ophthalmologist. The novelty of research is design of new models for prior diagnosis of diabetic retinopathy in patients for early precautionary measures. The models are rapid in classification for complex *** future;image dataset size can be increased to get more accurate results. © 2022

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Enhancing Local Descriptor Networks for Few-Shot Image Classification

Enhancing Local Descriptor Networks for Few-Shot Image Class...

引用

2023 International Annual Conference on Complex Systems and Intelligent science, CSIS-IAC 2023

作者： Dou, Quansheng Li, Xiujuan Tang, Huanling Liu, Yanchao Zhou, Liyuan College of Computer Science and Technology Shandong Technology and Business University Yantai China College of Computer Science and Technology Kashi University Xinjiang China College of Information and Electronic Engineering Shandong Technology and Business University Yantai China

ISBN: (纸本)9798350309003

Deep learning models rely on a large amounts of annotated data, and the performance of the model is unsatisfactory when the labeled samples are scarce. Existing few-shot learning methods based on local descriptors fail to consider correlation and usefulness between local descriptors. At the same time, due to the limited perceptual fields, it cannot make full use of the global information generated by distant pixels, leading to the problem of extracting features that are highly similar due to the same background, which in turn makes the model misclassify. In order to solve these problems, Fusioning Local and Global features Deep Nearest Neighbor Neural Network is proposed, which is gather relevant features through clusters of local descriptors and synthesize local and global information to learn more discriminative features, so as to further improve the classification performance of few-shot data. The experimental results in this paper show that the method shows superiority on miniImagNet, tieredImageNet and CUB-200-2011 datasets. © 2023 IEEE.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Hybrid Edge-Cloud Approach for Renewable Energy Management Using Deep Learning with Predictive Analytics

Hybrid Edge-Cloud Approach for Renewable Energy Management U...

引用

2024 IEEE International Conference on Recent Advances in science and engineering technology, ICRASET 2024

作者： Rekha, V. Sharmila, N. Komala, M. Pavithra, G.S. Mallikarjunaswamy, S. Naveen, K.B. Shilpa, M. Sheela, S. Department of Computer Science and Engineering Bangalore India Jss Science and Technology University Dept of Electronics and Electrical Engineering Mysuru India Sjb Institute of Technology Dept of Electronics and Communication Engineering Bengaluru India Jss Academy of Technical Eduction Dept of Computer Science and Engineering Bengaluru India Jss Academy of Technical Eduction Dept of Electronics and Communication Engineering Bengaluru India Bgs Institute of Technology Adichunchanagiri University Dept of Electronics and Electrical Engineering BG Nagara India Bengaluru Institute of Technology Dept of Information Science and Engineering Bengaluru India Global Academy of Technology Dept of Computer Sceinec and Engineering Bengaluru India

ISBN: (纸本)9798350388602

The integration of renewable energy sources into smart grids requires advanced management systems to optimize energy production, storage, and distribution. Conventional methods like Static Energy Allocation (SEA) and Rule-Based Control (RBC) struggle with real-time adaptability, leading to inefficiencies in handling the variable nature of renewable energy. SEA and RBC lack the predictive capabilities needed to effectively manage dynamic energy generation, resulting in increased energy losses and operational *** paper proposes a Hybrid Edge-Cloud Renewable Energy Management System (HECREMS) using Deep Learning with Predictive Analytics (DLPA). HECREMS integrates edge computing for real-time data processing with cloud computing for enhanced predictive analytics, providing a robust framework for efficient energy management. The proposed system leverages the low latency of edge devices and the computational power of the cloud to improve energy forecasting and decision-making. HECREMS outperforms traditional SEA and RBC methods, achieving a 0.25% improvement in energy utilization efficiency, a 0.20% reduction in operational costs, and a 0.30% increase in system responsiveness. By addressing the drawbacks of conventional approaches, HECREMS demonstrates significant advancements in the efficient and sustainable management of renewable energy within smart grids. © 2024 IEEE.

关键词： Cloud analytics

来源：评论

学校读者我要写书评

暂无评论

Blockchain Framework for Healthcare Data Management System in Clinical Trials 9

Blockchain Framework for Healthcare Data Management System i...

引用

9th International Conference on Fuzzy Systems and Data Mining, FSDM 2023

作者： Abdu, Nail Adeeb Ali Wang, Zhaoshun Abdualrhman, Mohammed Ahmed Ali Beijing10083 China University of Science and Technology Aden Yemen Department of Computer Science College of Engineering and Information Technology Al-Saeed University Taiz Yemen

ISBN: (纸本)9781643684703

The concept of information management involves the identification and organization of data into distinct sectors, sometimes referred to as silos, in order to anticipate and address future situations. One of the primary obstacles encountered in the administration of clinical trials pertains to the presence of information silos among the several stakeholders involved in the process. The improvement of cooperation and integration may be achieved by processing information in real-time and implementing a comprehensive system. This manuscript evaluates the viability of implementing the blockchain model for auditing clinical trials. It also proposes a blockchain-based solution that has the potential to enhance the overall quality of the practiced solutions and establish sustainable approaches for conducting clinical trials auditing. The suggested model focuses on the development of a grid-based system for organizing the many stages of a clinical trial. This involves creating distinct blocks for each phase and then constructing a complete network of these blocks to be shared across the consortium responsible for managing the trial data. The suggested architecture has substantial potential for enhancing security and adaptability, hence facilitating systematic improvements to information systems in the context of clinical trials. © 2023 The authors and IOS Press.

关键词： Blockchain

来源：评论

学校读者我要写书评

暂无评论

Algorithms of Sampling-Frequency-Independent Layers for Non-integer Strides 31

Algorithms of Sampling-Frequency-Independent Layers for Non-...

引用

31st European Signal Processing Conference, EUSIPCO 2023

作者： Imamura, Kanami Nakamura, Tomohiko Takamune, Norihiro Yatabe, Kohei Saruwatari, Hiroshi Graduate School of Information Science and Technology The University of Tokyo Tokyo Japan Department of Electrical Engineering and Computer Science Tokyo University of Agriculture and Technology Tokyo Japan

ISBN: (纸本)9789464593600

In this paper, we propose algorithms for handling non-integer strides in sampling-frequency-independent (SFI) convolutional and transposed convolutional layers. The SFI layers have been developed for handling various sampling frequencies (SFs) by a single neural network. They are replaceable with their non-SFI counterparts and can be introduced into various network architectures. However, they could not handle some specific configurations when combined with non-SFI layers. For example, an SFI extension of Conv-TasNet, a standard audio source separation model, cannot handle some pairs of trained and target SFs because the strides of the SFI layers become non-integers. This problem cannot be solved by simple rounding or signal resampling, resulting in the significant performance degradation. To overcome this problem, we propose algorithms for handling non-integer strides by using windowed sinc interpolation. The proposed algorithms realize the continuous-time representations of features using the interpolation and enable us to sample instants with the desired stride. Experimental results on music source separation showed that the proposed algorithms outperformed the rounding- and signal-resampling-based methods at SFs lower than the trained SF. © 2023 European Signal Processing Conference, EUSIPCO. All rights reserved.

关键词： Network architecture

来源：评论

学校读者我要写书评

暂无评论

Predictive Lung Disease Identification Using Artificial Intelligence and Machine Learning Algorithm 2

Predictive Lung Disease Identification Using Artificial Inte...

引用

2nd International Conference on Disruptive Technologies, ICDT 2024

作者： Singh, Rajani Das Lodhi, Abhinav Kumar Mishra, Amit Aeron, Anurag Sharma, Vyom Gl Bajaj Institute of Technology & Management Department of Information Technology UP Greater Noida India Raj Kumar Goel Institute of Technology Department of CSE-DS UP Ghaziabad India Graphic Era Hill University Computer Science & Engineering Uttarakhand Dehradun India Miet Computer Science & Engineering UP Meerut India Shri Ram Group of Colleges Computer Science & Engineering UP. Muzaffarnagar India

ISBN: (纸本)9798350371055

Lung sicknesses, which include continual obstructive pulmonary ailment (COPD), are a big international health difficulty, affecting hundreds of thousands of humans every 12 months. Traditional diagnostic strategies rely on subjective checks with the aid of medical professionals, leading to a high price of misdiagnosis and not on-time remedy. This highlights the need for an extra accurate and efficient lung ailment identification gadget. The Predictive Lung sickness identification (PLDI) gadget makes use of artificial intelligence (AI) and device-gaining knowledge of (ML) algorithms to locate and classify numerous lung diseases, including COPD. It uses techniques, image reputation, natural language processing, and records mining to research clinical snapshots and reviews, identifying patterns and markers of lung illnesses. The PLDI system has shown promising results in accurately detecting and classifying lung diseases, imparting a timely and correct diagnosis for patients. © 2024 IEEE.

关键词： Identification Misdiagnosis Obstructive Pulmonary Reputation Sicknesses

来源：评论

学校读者我要写书评

暂无评论

Leveraging Automated Methods and Advanced Neural Networks for Accurate information Extraction

Leveraging Automated Methods and Advanced Neural Networks fo...

引用

2023 IEEE International Conference on Paradigm Shift in information Technologies with Innovative Applications in Global Scenario, ICPSITIAGS 2023

作者： Hemalatha, S. Deshpande, Leena A. Shrimal, Gajendra Neetha, S.S. Alam, Intekhab Karthigaikumar, P. Prince Shri Venkateshwara Padmavathy Engineering College Department of Electrical and Electronics Engineering Chennai127 India Vishwakarma Institute of Information Technology Department of Computer Engineering Pune India Vivekananda Global University Department of Computer Science & Engineering Jaipur India Department of Computer Science and Information Technology Karnataka Bangalore India Maharishi School of Engineering and Technology Maharishi University of Information Technology Uttar Pradesh India Karpagam College of Engineering Department of Electronics and Communication Engineering Coimbatore641032 India

ISBN: (纸本)9798350329773

This research investigates how computerized techniques and superior neural networks may be used to improve the accuracy of fact extraction (IE). This study uses a deep learning model to extract principles and activities from unstructured texts. The experiment uses widespread IE benchmarks to evaluate the model's capability regarding idea and event extraction accuracy. Effects show that the proposed approach rides on the faculty of deep getting to know, leveraging technique capabilities with the herbal language era. It allows for the progressed accuracy of IE over the modern-day baseline strategies. The model also achieves a precision price of 76.79% in concept extraction, with a keep-in-mind charge of 79.92%. In event extraction, the gadget reports a precision fee of 64.17% and a recall rate of 69.66%. The proposed technique is expected to be used to efficiently enhance the overall performance of IE tasks in a wide variety of programs. © 2023 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Predictive Cloud-Based Application for the Early Detection and Risk Stratification of Oral Cancer using Keratin Pearls, Multiple Nucleoli and Vascular Invasion 28

Predictive Cloud-Based Application for the Early Detection a...

引用

28th International computer science and engineering Conference, ICSEC 2024

作者： Benagi, Abhinav Narayan, Dhanyatha Priya, K. Shetty, Savita K Augustine, Dominic Sv, Sowmya Kanavalli, Anita Rage, Charith Reddy, A Susmitha Ramaiah Institute of Technology Dept of Computer Science and Engineering Bangalore India Ramaiah Institute of Technology Dept of Information Science and Engineering Bangalore India MS Ramaiah Dental College and Hospital Dept of Oral Pathology and Microbiology Bangalore India

ISBN: (纸本)9798350366860

Oral cancer is a significant public health concern, with high morbidity and mortality rates. Early detection is critical for improving patient outcomes and reducing the burden of the disease. Histopathological images, essential for distinguishing between normal, oral precancerous, and cancerous lesions, may miss critical diagnostic features of aggressiveness. Additionally, routine histopathology struggles to effectively identify high-risk OSCC patients. This work was conducted in collaboration with Ramaiah Institute of technology and Ramaiah Dental College to improve OSCC detection through a pipeline-based computer-Assisted screening approach involving classification, feature detection, and risk stratification. First, the classification module achieved an accuracy of 99.25% with EfficientNet in classifying cancerous and non-cancerous lesions. The feature detection module successfully identified key indicators of aggressiveness, including keratin pearls and multiple nucleoli using YOLO-NAS, while vascular invasion was detected by the EfficientNet classifier. Finally, we stratified cancer risk based on these detected features, producing a comprehensive pathology report that provides clinicians with valuable insights for informed diagnosis. Ultimately, this work aims to provide doctors with insights from analyses using models trained on key features like keratin pearls, multiple nucleoli, and vascular invasion. © 2024 IEEE.

关键词： Lung cancer

来源：评论

学校读者我要写书评

暂无评论

Enhancing Tumor Detection through ANFIS-Based Clustering in Biomedical Imaging 2

Enhancing Tumor Detection through ANFIS-Based Clustering in ...

引用

2nd International Conference on Computational Intelligence and Network Systems, CINS 2024

作者： Sri, Sweta P. Lakshay, Morwal Suguna, M. Department of Information Technology Sri Sivasubramaniya Nadar College of Engineering Tamil Nadu Chennai603110 India School of Computer Science and Engineering Vellore Institute of Technology Tamil Nadu Chennai600127 India

ISBN: (纸本)9798331504106

In Biomedicine, computer software is applied to the medical image for content analysis by MRI scanning for the program. Preprocessing, removal of noise, processing images in order to detect objects and analysis of the images for tumor assessment in terms of shape and volume form part of the MRI based brain tumor detection process. The author of this paper presents a method employing ANFIS models with clustering in order to determine the number of clusters necessary for identifying brain and lymph node tumors. ANFIS improves classification as well as cancer diagnosis and therapy by automating selection and refinement of treated clusters. ©2024 IEEE.

关键词： Medical image processing

来源：评论

学校读者我要写书评

暂无评论

Emotion Classification and Respective Sticker Mapping Using Convolutional Neural Networks 14

Emotion Classification and Respective Sticker Mapping Using ...

引用

14th International Conference on Computing Communication and Networking Technologies, ICCCNT 2023

作者： Ghosh, Sayan Karmakar, Raja Chatterjee, Twinkle Ghosal, Sandipan Techno International New Town Computer Science and Engineering Kolkata700156 India Heritage Institute of Technology Computer Science and Engineering Kolkata700107 India Techno International New Town Information Technology Kolkata700156 India

ISBN: (纸本)9798350335095

The utilization of social media is aggravating and so is the demand of using messaging apps, to stay connected. People now not only send text messages but also prefer using stickers which is the new alternative to emojis. While an emoji can only portray emotion, stickers are used to graphically illustrate and capture the human mind's attention by being more relatable and accurate. One thing to consider here is that men and women have a different choice of emoji to express their emotion so a random emoji may be ambiguous, i.e., confusing to the other gender. Now, emotion analysis has gained significant importance through online media in recent years. Thus, we have tried to primarily focus on two steps, namely emotion classification and gender analysis for sticker generation. Therefore, this paper employs machine learning models to classify emotions and predict gender to generate accurate stickers. Convolutional Neural Networks (CNNs) are highly productive in feature extraction and object classification on images which is why we have used a CNN for emotion analysis and another CNN for gender analysis combining both we can map the stickers from a prebuilt directory. © 2023 IEEE.

关键词： Convolutional neural networks

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：