检索结果-内蒙古大学图书馆

2nd International Conference on Electronics and Renewable System., ICEARS 2023

作者： Jose Triny, K. Kesavan, S. Navadeepan, H. Ranjith Kum.r, P. M. Kumarasamy College of Engineering Department of Computer Science and Engineering Karur India

ISBN: (纸本)9798350346640

Scientists can now assess the expression of hundreds of genes concurrently in a single experim.nt because to the advancem.nts in DNA m.croarray technology. Gene expression patterns, which depict the m.lecular state of a cell, offer im.ense potential as a diagnostic tool. The solutions to resolving the fundam.ntal harm. associated with diagnosis and discovery is known to be found in the categorization of diseases transcriptional activation utilizing data. Com.lete sim.ltaneous m.nitoring of several gene expressions is now achievable because to the recent developm.nt of the DNA m.croarray technology. The possibility of disease diagnosis utilizing gene expression data has begun to be explored by specialists thanks to the abundance of gene expression data. The previous years, a sizable num.er of procedures there have planned with positive outcom.s. Em.edded techniques have the prevalence advantage of interacting with the classification algorithm.;while on the sam. tim. have decrease com.utational tim. com.ared to wrapper strategies. A deeper exam.nation of the issue, the suggested rem.dies, and the related problem. as a whole is required to develop understanding of the disease categorization challenge. In this study, we present a thorough searching m.thod, clustering m.thod, as well as classification techniques including such Trend sim.larity such as algorithm.for particle swarm.optim.zation and evaluate them.predicated on their review tim., classification accuracy, and capacity to reveal biologically significant gene inform.tion. Depending on our classification m.thod, we diagnose conditions including cancer (of the lung, heart, breast, and skin) and other conditions. We also determ.ne the severity of conditions and recom.end m.dications for those conditions using deep learning algorithm.. Our test results dem.nstrate better accuracy in classifier perform.nce when used with graphs. © 2023 IEEE.

关键词： Diseases

来源：评论

学校读者我要写书评

暂无评论

Distance based Sim.larity Search with Distinct Encrypted Im.ge Storage on Cloud 2

Distance based Similarity Search with Distinct Encrypted Ima...

引用

2nd International Conference on Electronics and Renewable System., ICEARS 2023

作者： m.ni, V. Akash, K. m.ham.d Sulaim.n, S. M. Kumarasamy College of Engineering Department of Computer Science and Engineering Karur India

ISBN: (纸本)9798350346640

Secure picture recovery just on cloud has becom. increasingly popular in recent years as a result of the introduction of cloud com.uting. Both businesses and people frequently use the internet to m.nage and store their sensitive inform.tion, including such old photos and personal health records, due to its high level of convenience and financial savings. The typical m.thod would be to encrypt the inform.tion prior to it being exported to using cloud to guarantee the secrecy of the inform.tion. The retrieval of inform.tion from.encrypted data, for instance, m.y be im.ossible due to conventional encryption. The difficult issue in the cipher - text scenario is how to accom.lish an effective retrieval while m.intaining client anonym.ty. This study suggests a cutting-edge im.ge retrieval m.thod that achieves excellent efficiency and confidentiality over encrypted cloud data. An indexing tree is frequently used in the inform.tion extraction strategy to increase search perform.nce. The security of private cloud data, including outsourced photos, the search tree, and query requests, is also a m.jor concern. First, use an algorithm.for extracting features from.integrated picture features, which are m.de up of fundam.ntal aspects like colour and shape. In particular, proposed technique can reach logarithm.c search tim. thanks to the use of a balancing index tree. Second, the picture and query feature are encrypted using the secure inner product. Additionally, put in place a m.thod to detect duplicate im.ge m.terial. Finally, extract the data owner details based on uploaded user-subm.tted im.ge attributes that have been linked with distance m.asurem.nts. © 2023 IEEE.

关键词： Cryptography

来源：评论

学校读者我要写书评

暂无评论

Deep Reinforcem.nt Learning with Graph Neural Networks: An Indepth Analysis of Algorithm. and Applications 2

Deep Reinforcement Learning with Graph Neural Networks: An I...

引用

2nd International Conference on Advanced Com.uting and Com.unication Technologies, ICACCTech 2024

作者： Srivastava, Sakshi m.ttal, Harish Kum.r Vasesi, Sonika Param.eet Jain, Avdhi Chauhan, Divya B M Institute of Engineering and Technology Computer Science and Engineering Sonipat India

ISBN: (纸本)9798331519056

Deep Reinforcem.nt Learning (DRL) enable several areas of artificial intelligence, including perception recognition, expert system. recom.ender program.and gam.. Also, graph neural networks (GNN) have shown excellent perform.nce in supervised m.chine learning of graph structure data. Recently, com.ining GNN and DRL in graphical arranged environm.nts has captivated m.ch attention. This article provides a panoram.c overview of all hybrid work. In order to facilitate generalisation, we incorporate Graph Neural Networks (GNN) into DRL agents in this research and create an action space that is problem.specific. Deep Learning m.dels called GNNs are innately m.de to generalise across graphs with varying sizes and architectures. This enables a broad range of network designs to ne learnt by proposed GNN-based DRL operator. © 2024 IEEE.

关键词： Deep Learning Deep Reinforcem.nt Graph Neural System.Network Hybrid DRL -GNN Query

来源：评论

学校读者我要写书评

暂无评论

Optim.zing Energy Efficiency and Security in Wireless Sensor Networks with a Hybrid HF-ECC 10

Optimizing Energy Efficiency and Security in Wireless Sensor...

引用

10th International Conference on Advanced Com.uting and Com.unication System., ICACCS 2024

作者： Tharani, B. Devi, B. Padm.ni M. Kumarasamy College of Engineering Department of Computer Science Tamil Nadu Thalavapalayam639113 India

ISBN: (纸本)9798350384369

Num.rous distinct categories of businesses take advantage of 'Wireless Sensor Networks (WSNs)'. Interfering with 'Sensor Node (SN s)' and eavesdropping on com.unications are two possible attacks. Physically placed SN s provide an additional opportunity for hackers to alter the inform.tion being transm.tted or establish unauthorized connections over the WSN network. Verifying the route and encrypting inform.tion transm.tted across connected SN s are a couple of steps toward WSN privacy. The network's SN s m.st additionally possess an authentication key for encrypting and decrypting processes. As a result, proper key adm.nistration is crucial for WSN security. Secured and energy-effective transfer of inform.tion within WSN needs to obtain suitable key organizations in environm.nts with lim.ted resources. There are several security concerns with existing 'Elliptic Curve Cryptography (ECC)' techniques, including their cost of com.utation and susceptibility to m.ltiple assaults. A 'Hash- Function (HF)' relied on m.C authenticating protocol and an ECC built around 'Secrecy Connectivity Probability (SCP)' is proposed in this paper as a hybrid (HF-ECC) protocol. The originating SN uses the HF for authenticating against its neighboring SN s using the m.C Key, which produces pre- distribution keys, once the SNs are deployed throughout the WSN region. For determ.ning the key, each SN needs to perform.broadcasting with its identity;this will get them.accessibility to the self-organized through a secure process. Upon route establishm.nt, the ECC began working in the origin SN for encryption of the inform.tion ahead of transm.ssion. Whenever the m.ssage that was encrypted arrives at the recipient SN, the ECC decrypts it. Consequently, the proposed HF-ECC protocol self-organizes to offer a secure accessibility process for every SN within the WSN. In com.arison to the existing ECC, the proposed hybrid (HF-ECC) protocol outperform. it regarding both security and energy efficiency, a

关键词： Hash functions

来源：评论

学校读者我要写书评

暂无评论

Rear Guard: LCD-based Accident Prevention System.for Large Vehicle 5

Rear Guard: LCD-based Accident Prevention System for Large V...

引用

5th International Conference on IoT Based Control Networks and Intelligent System., ICICNIS 2024

作者： Ram.a Shri, G. Aathika Erfana, S. Devadharshini, A. Harinika, A. Keerthana, G. M. Kumarasamy College of Engineering Thalavapalayam Department of Computer Science and Engineering Tamil Nadu Karur639113 India

ISBN: (数字)9798331518097

ISBN: (纸本)9798331518097

This project deals with issues and risks that the drivers face especially while overtaking big vehicles in the way. While facing this challenge, a solution aim.d at enhancing the safety on roads. Our technology uses a wireless cam.ra m.unted at the front side of the autom.bile and transm.ts live feed onto a screen at the rear side of the vehicle and achieves the see-through effect. This would ensure the passing of the car with a view to oncom.ng traffic, thus reducing accidents and collisions through sudden braking or unsafe overtaking. This project designed a prototype in which the vehicle is controlled through Bluetooth using an Android application for controlling a vehicle. The four gear m.tors connected to the m.crocontroller on the receiver side enable the vehicle to m.ve. © 2024 IEEE.

关键词： Liquid crystal displays

来源：评论

学校读者我要写书评

暂无评论

An IoT Based Fram.work for m.nitoring Power Consum.tion in Sm.rt Grids

An IoT Based Framework for Monitoring Power Consumption in S...

引用

2024 International Conference on Com.unication, computer sciences and engineering, IC3SE 2024

作者： Raghavendra, R. Chauhan, Nidhi m.ttal, Gaurav Department Of Computer Science And Information Technology Bangalore India Noida Institute Of Engineering And Technology Department Of Information Technology & M. Tech Integrated Uttar Pradesh Greater Noida India Tula's Institute Department Of Computer Science And Engineering Dehradun India

ISBN: (纸本)9798350366846

The electrical power system.grid in India, com.only referred to as the power grid, has been in operation for an extensive duration. During this tim. period, there were no significant advancem.nts or alterations docum.nted in the electricity grid infrastructure. The need for electrical power consum.tion is seeing a significant increase, but the current grid infrastructure is unable to m.et these growing dem.nds. In order to m.et the dem.nds of future power load, it is im.erative to develop an enhanced system.that possesses the qualities of reliability, security, intelligence, and efficiency. This study develops a power m.nagem.nt system.and explains its autom.tion and functionality by utilising sm.rt grid and IoT technologies. The research suggests a m.thod for keeping an eye on consum.r power use and controlling the use of electrical devices by putting in place a system.that can identify voltage, current, and weather conditions in addition to alerting users in advance of possible problem. with power usage. Future system.im.rovem.nts that are region-specific and intended to reduce loads and m.xim.se system.perform.nce are m.de possible by the data collected. © 2024 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Autom.ted Sem.ntic Interpretations in Web Applications via Cloud Com.uting Aspects

Automated Semantic Interpretations in Web Applications via C...

引用

2024 International Conference on Com.unication, computer sciences and engineering, IC3SE 2024

作者： Preethi, D. m.hta, Poornim. Sood, Devendra Department Of Computer Science And Information Technology Bangalore India Noida Institute Of Engineering And Technology Department Of It And M. Tech Integrated Uttar Pradesh Greater Noida India Tula's Institute Department Of Computer Science And Engineering Dehradun India

ISBN: (纸本)9798350366846

The creation and execution of applications in several disciplines have been greatly aided by the integration of cloud com.uting and sem.ntic web technologies. Although they are independent entities, they m.y also be com.ined in m.ny ways to provide answers - a subject that has been thoroughly researched lately. The num.er of 'new cloud providers, applications, facilities, m.nagem.nt system., and data' has significantly increased in recent years. The com.lexity that has resulted from.this expansion em.hasises the need for new technology in order to efficiently m.nage and support these varied and m.ny services and resources. As a result, difficulties with 'cloud services and resources' transferability, com.atibility, security, choice, negotiation, identity, and specification m.ght arise. Sem.ntic Technologies are a vital tool for reassessing these issues because of their enorm.us 'cloud com.uting' capacity. This study uses a variety of sources to exam.ne and evaluate the im.act of 'Sem.ntic-Web technology' in the cloud. m.reover, a cloud-based engagem.nt strategy shows how we m.y leverage the features and advantages of cloud com.uting to build the sem.ntic web and provide autom.tic sem.ntic annotations to online applications on a large scale. © 2024 IEEE.

关键词： Sem.ntic Web

来源：评论

学校读者我要写书评

暂无评论

Class-conditional dom.in adaptation for sem.ntic segm.ntation

引用

Com.utational Visual m.dia 2024年第5期10卷 1013-1030页

作者： Yue Wang Yuke Li Jam.s H.Elder Runm.n Wu Huchuan Lu School of Information and Communication Engineering Dalian University of TechnologyDalian 116024China School of Computer Science Wuhan UniversityWuhan 430072China Department of Electrical Engineering and Computer Science York UniversityToronto M3J 1P3Canada Department of Computer Science the University of Hong KongHong Kong 999077China

Sem.ntic segm.ntation is an im.ortant sub-task for m.ny ***,pixel-level ground-truth labeling is costly,and there is a tendency to overfit to training data,thereby lim.ting the generalization *** dom.in adaptation can potentially address these problem. by allowing system. trained on labelled datasets from.the source dom.in(including less expensive synthetic dom.in)to be adapted to a novel target *** conventional approach involves autom.tic extraction and alignm.nt of the representations of source and target dom.ins *** lim.tation of this approach is that it tends to neglect the differences between classes:representations of certain classes can be m.re easily extracted and aligned between the source and target dom.ins than others,lim.ting the adaptation over all ***,we address:this problem.by introducing a Class-Conditional Dom.in Adaptation(CCDA)*** incorporates a class-conditional m.lti-scale discrim.nator and class-conditional losses for both segm.ntation and ***,they m.asure the segm.ntation,shift the dom.in in a classconditional m.nner,and equalize the loss over *** results dem.nstrate that the perform.nce of our CCDA m.thod m.tches,and in som. cases,surpasses that of state-of-the-art m.thods.

关键词： dom.in adaptation generative adversarial networks sem.ntic segm.ntation cityscapes

来源：评论

学校读者我要写书评

暂无评论

Lightweight Cloud-Based Solution for Digital Education and Assessm.nt

引用

science and technology Libraries 2024年第3期43卷 274-286页

作者： Geetha, J. Jayalakshm., D.S. Naresh, E. Sreenivasa, N. Department of Computer Science & Engineering M. S. Ramaiah Institute of Technology Affiliated to VTU Bengaluru India Department of Information Technology Manipal Institute of Technology Bengaluru Manipal Academy of Higher Education Manipal India Department of CSE Nitte Meenakshi Institute of Technology Bangalore India

With a shift from.the regular in-class education to e-learning-based education, there is a need to have the right platform.to conduct assessm.nts and courses in a m.re effective and user-friendly m.nner which includes robust m.nitoring, capability of doubt clearing, easy sharing of reference m.terials and m.re. There is a need for a highly custom.zable lab which supports all languages and is very lightweight. Our solution focuses on the creation of a robust, lightweight lab that provides a plug-and-play m.del for on-dem.nd com.ilation capabilities. This virtual lab is hosted on the in-house blade servers and virtualized using m.tal-as-a-service (m.as) along with juju controller. Kubernetes is used for on-dem.nd com.ilation support as well as serves as the load balancer for the servers. The lab is specific to computer science-related courses and aim. to im.act the students present in the low bandwidth and rural areas. m.reover, the proposed m.thodology enhances the capabilities of the institutions to utilize their resources to the fullest with the help of virtualization through m.as, juju and open-stack and orchestration through Kubernetes. © 2023 The Author(s). Published with license by Taylor & Francis Group, LLC.

关键词： Laboratories

来源：评论

学校读者我要写书评

暂无评论

Im.EDRD from.Retinal Fundus Im.ges Using m.lti-Level Classification Techniques

引用

Intelligent Autom.tion & Soft Com.uting 2023年第1期35卷 567-580页

作者： m.P.Karthikeyan E.A.m.ry Anita Department of Computer Science and Engineering R.M.K.Engineering CollegeChennai601206India Department of Computer Science and Engineering School of Engineering and TechnologyChrist UniversityBengaluru560029India

In recent years,there has been a significant increase in the num.er of people suffering from.eye illnesses,which should be treated as soon as possible in order to avoid *** Fundus im.ges are em.loyed for this purpose,as well as for analysing eye abnorm.lities and diagnosing eye *** can be recognised as bright lesions in fundus pictures,which can be thefirst indicator of diabetic *** that in m.nd,the purpose of this work is to create an Integrated m.del for Exudate and Diabetic Retinopathy Diagnosis(Im.EDRD)with m.lti-level *** m.del uses Support Vector m.chine(SVm.-based classification to separate norm.l and abnorm.l fundus im.ges at thefirst *** input pictures for SVm.are pre-processed with Green Channel Extraction and the retrieved features are based on Gray Level Co-occurrence m.trix(GLCm..Furtherm.re,the presence of Exudate and Diabetic Retinopathy(DR)in fundus im.ges is detected using the Adaptive Neuro Fuzzy Inference System.ANFIS)classifier at the second level of *** detection,blood vessel extraction,and Optic Disc(OD)detection are all processed to achieve suitable ***,the second level processing com.rises m.rphological Com.onent Analysis(m.A)based im.ge enhancem.nt and object segm.ntation processes,as well as feature extraction for training the ANFIS classifier,to reliably diagnose ***,thefindings reveal that the proposed m.del surpasses existing m.dels in term. of accuracy,tim. efficiency,and precision rate with the lowest possible error rate.

关键词： Retinal fundus im.ges exudate diabetic retinopathy SVm ANFIS m.rphological com.onent analysis

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：