检索结果-南通市图书馆

南京医科大学学报（自然科学版） 2007年第4期27卷 1122-1122页

作者： Qiu, j. Ni, Y. H. Gong, H. x. Fei, L. Pan, x. Q. guo, M. Chen, r. H. guo, x. r. Nanjing Med Univ Nanjing Maternal ＆ Child Hlth Hosp Dept Pediat Nanjing 210029 Peoples R China. Nanjing Med Univ Dept Pediat Nanjing 210029 Peoples R China.

Identification of differ.ntially ex.r.ssed genes in omental adipose tissues of obese patients by suppr.ssion subtr.ctive hybr.dization

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

南京医科大学学报（自然科学版） 2007年第5期27卷 427-427页

作者： Qiu, J. Ni, Y. H. Gong, H. x. Fei, L. Pan, x. Q. guo, M. Chen, r. H. guo, x. r. Nanjing Med Univ Nanjing Maternal ＆ Child Hlth Hosp Dept Pediat Nanjing 210029 Peoples R China Nanjing Med Univ Dept Pediat Nanjing 210029 Peoples R China.

To identify differ.ntially ex.r.ssed genes between obese individuals and nor.al contr.l, we have under.aken suppr.ssion subtr.ctive hybr.dization (SSH)。Omental adipose tissues wer. obtained via abdominal sur.er. for.appendicitis in both 13 obese subjects[BMI (body mass index. > 30 kg/m(2)] and 13 nor.al subjects (BMI > 18 and < 25 kg/m(2))。

关键词：肥胖症脂肪组织削减杂交差异表达基因葡萄糖转运泵

r.bust Deep Image Water.ar.ing: A Sur.ey

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computer., Mater.als and Continua 2024年第1期81卷 133-160页

作者： Luo Y. Tan x. Cai Z. College of Computer National University of Defense Technology Changsha 410005 China

In the er. of inter.et pr.lifer.tion, safeguar.ing digital media copyr.ght and integr.ty, especially for.images, is imper.tive. Digital water.ar.ing stands out as a pivotal solution for.image secur.ty. With the advent of deep lear.ing, water.ar.ing has seen significant advancements. Our.r.view focuses on the innovative deep water.ar.ing appr.aches that employ neur.l networ.s to identify r.bust embedding spaces, r.silient to var.ous attacks. These methods, char.cter.zed by a str.amlined encoder.decoder.ar.hitectur., have shown enhanced per.or.ance thr.ugh the incor.or.tion of novel tr.ining modules. This ar.icle offer. an in-depth analysis of deep water.ar.ing’s cor. technologies, cur.ent status, and pr.spective tr.jector.es, evaluating r.cent scholar.y contr.butions acr.ss diver.e fr.mewor.s. It concludes with an over.iew of the technical hur.les and pr.spects, pr.viding essential insights for.ongoing and futur. r.sear.h endeavor. in digital image water.ar.ing. © 2024 The Author.. Published by Tech Science Pr.ss.

关键词： data pr.tection deep neur.l networ. Deep image water.ar.ing multimedia secur.ty

Hier.r.hical OptimizationMethod for.Feder.ted Lear.ing with Featur. Alignment and Decision Fusion

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computer., Mater.als and Continua 2024年第1期81卷 1391-1407页

作者： Li K. Wang x. Wang H. College of Computer.Science and Engineer.ng North Minzu University Yinchuan 750021 China The Key Labor.tor. of Images & Gr.phics Intelligent Pr.cessing of State Ethnic Affair. Commission North Minzu University Yinchuan 750021 China

In the r.alm of data pr.vacy pr.tection, feder.ted lear.ing aims to collabor.tively tr.in a global model. However. heter.geneous data between clients pr.sents challenges, often r.sulting in slow conver.ence and inadequate accur.cy of the *** shar.d featur. r.pr.sentations alongside customized classifier. for.individual clients emer.es as a pr.mising per.onalized solution. Nonetheless, pr.vious r.sear.h has fr.quently neglected the integr.tion of global knowledge into local r.pr.sentation lear.ing and the syner.y between global and local classifier., ther.by limiting model per.or.ance. To tackle these issues, this study pr.poses a hier.r.hical optimization method for.feder.ted lear.ing with featur. alignment and the fusion of classification decisions (FedFCD). FedFCD r.gular.zes the r.lationship between global and local featur. r.pr.sentations to achieve alignment and incor.or.tes decision infor.ation fr.m the global classifier. facilitating the late fusion of decision outputs fr.m both global and local classifier.. Additionally, FedFCD employs a hier.r.hical optimization str.tegy to flex.bly optimizemodel par.meter.. Thr.ugh ex.er.ments on the Fashion-MNIST, CIFAr.10 and CIFAr.100 datasets, we demonstr.te the effectiveness and super.or.ty of FedFCD. For.instance, on the CIFAr.100 dataset, FedFCD ex.ibited a significant impr.vement in aver.ge test accur.cy by 6.83% compar.d to four.outstanding per.onalized feder.ted lear.ing appr.aches. Fur.her.or., ex.ended ex.er.ments confir.the r.bustness of FedFCD acr.ss var.ous hyper.ar.meter.values. © 2024 The Author.. Published by Tech Science Pr.ss.

关键词： data heter.geneity decision fusion featur. alignment Feder.ted lear.ing hier.r.hical optimization

Epidur.l Analgesia in the Latent Phase of Labor.and the r.sk of Cesar.an Deliver. A Five-year.r.ndomized Contr.lled Tr.al

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

南京医科大学学报（自然科学版） 2009年第11期29卷 1628-1628页

作者： Wang, F. Z. Shen, x. F. guo, x. r. Peng, Y. Z. Gu, x.Q. Nanjing Med Univ Dept A nesthesiol Affiliated Nanjing Matern & Child Hlth Care Hosp Nanjing 210004 Jiangsu Peoples R China 不详

Population Dynamics and Key Influencing Factor. of Micr.thr.x.par.icella Caused Sludge Bulking

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Huanjing Kex.e/Envir.nmental Science 2024年第11期45卷 6766-6774页

作者： Su G.-x. Qi r. x.ao S.-M. Dou D.-C. guo Y.-x. Gao J. Tianjin Key Labor.tor. of Aquatic Science and Technology School of Environmental and Municipal Engineering Tianjin Chengjian University Tianjin 300384 China State Key Labor.tor. of Envir.nmental Aquatic Chemistr. Research Center for Eco-Environmental Sciences Chinese Academy of Sciences Beijing 100085 China

Given the fr.quent occur.ence of the sludge bulking phenomena caused by Micr.thr.x.par.icella（M. par.icella）over.r.wth dur.ng the low-temper.tur. per.od in the full-scale municipal wastewater.tr.atment plant，the filamentous bacter.a dynamics of the M. par.icella population wer. systematically investigated dur.ng the completed bulking per.od using the multi-dimensional integr.ted filamentous monitor.ng method combined with the molecular.quantitative method. A lab-scale r.actor.was constr.cted and successfully simulated bulking phenomena under.the low-temper.tur. condition. The tr.nd of dominant filamentous bacter.al shift and the main influencing par.meter. wer. estimated dur.ng the bulking per.od，to ver.fy the key factor. for.filament gr.wth and ex.inction of the sludge bulking phenomenon in the full-scale plant. The r.sults showed that the bulking-caused M. par.icella was char.cter.zed by significant changes in the diver.ity of filamentous bacter.a，and the dominant filamentous bacter.al succession was r.flected in the competition between the skeleton filamentous bacter.a Type 0092 and Type 0041 and M. par.icella. At lower.temper.tur.s，sludge loading was an impor.ant contr.lling factor.for.the competitive gr.wth advantage of M. par.icella：when the temper.tur. was lower.appr.x.mately 15°C），low sludge loading［0.05 kg·（kg·d）−1］could lead to the stabilized elevation of M. par.icella，and with the loading incr.ased to 0.1 kg·（kg·d）−1 or.mor.，M. par.icella populations showed a tar.eted decline despite the pr.sence of the low temper.tur. factor. © 2024 Science Pr.ss. All r.ghts r.ser.ed.

关键词： filamentous bacter.a Micr.thr.x.par.icella population dynamics sludge bulking sludge loading

NCCMF: Non-Collabor.tive ContinuousMonitor.ng Fr.mewor. for.Container.Based Cloud r.ntime Status

同方期刊数据库评论

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Computer., Mater.als and Continua 2024年第1期81卷 1687-1701页

作者： Zheng T. Tang W. Chen x. Shen C. School of Cyber.Science and Engineer.ng Sichuan University Chengdu 610065 China Intelligent Policing and National Secur.ty r.sk Management Labor.tor. Luzhou 646000 China Key Labor.tor. of Data Pr.tection and Intelligent Management (Sichuan Univer.ity) Ministry of Education Chengdu 610065 China Cyber.Science r.sear.h Institute Sichuan University Chengdu 610065 China

The secur.ty per.or.ance of cloud ser.ices is a key factor.influencing user.' selection of Cloud Ser.ice Pr.vider. (CSPs). Continuous monitor.ng of the secur.ty status of cloud ser.ices is cr.tical. However. ex.sting r.sear.h lacks a pr.ctical fr.mewor. for.such ongoing monitor.ng. To addr.ss this gap, this paper.pr.poses the fir.t Non- Collabor.tive Container.Based Cloud Ser.ice Oper.tion State Continuous Monitor.ng Fr.mewor. (NCCMF), based on r.levant standar.s. NCCMF oper.tes without the CSP's collabor.tion by: 1) establishing a scalable super.isor. index.system thr.ugh the identification of secur.ty r.sponsibilities for.each r.le, and 2) designing a Continuous Metr.cs Super.ision Pr.tocol (CMA) to automate the negotiation of super.isor. metr.cs. The fr.mewor. also outlines the super.ision pr.cess for.cloud ser.ices acr.ss differ.nt deployment models. Ex.er.mental r.sults demonstr.te that NCCMF effectively monitor. the oper.tional state of two r.al-wor.d IoT (Inter.et of Things) cloud ser.ices, with an aver.ge super.ision er.or.of less than 15%. © 2024 The Author.. Published by Tech Science Pr.ss.

关键词： Container.based cloud continuous monitor non-collabor.tive r.ntime status

Distr.buted field mapping for.mobile sensor.teams using a der.vative‐fr.e optimisation algor.thm

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

IET Cyber.Systems and r.botics 2024年第2期6卷 20-34页

作者： Tony x.Lin Jia guo Said Al-Abr. Fumin Zhang Fumin Zhang31School of Electr.cal&Computer.Engineer.ng Georgia Institute of TechnologyAtlantaGeorgiaUSA College of Engineer.ng Sultan Qaboos UniversityMuscatOman Depar.ment of Electr.nic&Computer.Engineer.ng Hong Kong University of Science and TechnologyHong KongChina

The author. pr.pose a distr.buted field mapping algor.thm that dr.ves a team of r.bots to ex.lor. and lear. an unknown scalar.field using a Gaussian Pr.cess(GP).The author.’str.tegy ar.ses by balancing ex.lor.tion objectives between ar.as of high er.or.and high *** computing high er.or.r.gions is impossible since the scalar.field is unknown,a bio-inspir.d appr.ach known as Speeding-Up and Slowing-Down is lever.ged to tr.ck the gr.dient of the GP *** appr.ach achieves global field-lear.ing conver.ence and is shown to be r.sistant to poor.hyper.ar.meter.tuning of the *** appr.ach is validated in simulations and ex.er.ments using 2D wheeled r.bots and 2D flying mini-atur. autonomous blimps.

关键词： envir.nment sensing multi-r.bot systems

r.-Distr.buting Facial Featur.s for.Engagement Pr.diction with Moder.TCN

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computer., Mater.als and Continua 2024年第1期81卷 369-391页

作者： Li x. Zhu W. Li Q. Hou C. Zhang Y. College of Infor.ation and Ar.ificial Intelligence Nanchang Institute of Science and Technology Nanchang 330108 China School of Electr.cal and Infor.ation Engineer.ng Wuhan Institute of Technology Wuhan 430205 China School of Electr.nic Infor.ation Engineer.ng Wuhan Donghu University Wuhan 430212 China

Automatically detecting lear.er.’ engagement levels helps to develop mor. effective online teaching and assessment pr.gr.ms, allowing teacher. to pr.vide timely feedback and make per.onalized adjustments based on students’ needs to enhance teaching effectiveness. Tr.ditional appr.aches mainly r.ly on single-fr.me multimodal facial spatial infor.ation, neglecting tempor.l emotional and behaviour.l featur.s, with accur.cy affected by significant pose var.ations. Additionally, convolutional padding can er.de featur. maps, affecting featur. ex.r.ction’s r.pr.sentational capacity. To addr.ss these issues, we pr.pose a hybr.d neur.l networ. ar.hitectur., the r.distr.buting facial featur.s and tempor.l convolutional networ. (r.fEIP). This networ. consists of thr.e key components: fir.t, utilizing the spatial attention mechanism lar.e ker.el attention (LKA) to automatically captur. local patches and mitigate the effects of pose var.ations; second, employing the featur. or.anization and weight distr.bution (FOWD) module to r.distr.bute featur. weights and eliminate the impact of white featur.s and enhancing r.pr.sentation in facial featur. maps. Finally, we analyse the tempor.l changes in video fr.mes thr.ugh the moder. tempor.l convolutional networ. (Moder.TCN) module to detect engagement levels. We constr.cted a near.infr.r.d engagement video dataset (NEVD) to better.validate the efficiency of the r.fEIP networ.. Thr.ugh ex.ensive ex.er.ments and in-depth studies, we evaluated these methods on the NEVD and the Database for.Affect in Situations of Elicitation (DAiSEE), achieving an accur.cy of 90.8% on NEVD and 61.2% on DAiSEE in the four.class classification task, indicating significant advantages in addr.ssing engagement video analysis pr.blems. © 2024 The Author.. Published by Tech Science Pr.ss.

关键词： Engagement pr.diction r.-distr.buting facial featur.s spatiotempor.l networ. tempor.l convolutional networ.

Digital Image Steganogr.pher.Identification: A Compr.hensive Sur.ey

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computer., Mater.als and Continua 2024年第1期81卷 105-131页

作者： Zhang Q. Zhang Y. Ma Y. Liu Y. Luo x. Infor.ation Engineer.ng Univer.ity Zhengzhou 450001 China Key Labor.tor. of Cyber.pace Situation Awar.ness of Henan Pr.vince Zhengzhou 450001 China College of Computer.and Infor.ation Engineer.ng Henan Normal University Xinxiang 453007 China

The r.pid development of the inter.et and digital media has pr.vided convenience while also posing a potential r.sk of steganogr.phy abuse. Identifying steganogr.pher.is essential in tr.cing secr.t infor.ation or.gins and pr.venting illicit cover. communication online. Accur.tely discer.ing a steganogr.pher.fr.m many nor.al user. is challenging due to var.ous factor., such as the complex.ty in obtaining the steganogr.phy algor.thm, ex.r.cting highly separ.bility featur.s, and modeling the cover.data. After.ex.ensive ex.lor.tion, sever.l methods have been pr.posed for.steganogr.pher.identification. This paper.pr.sents a sur.ey of ex.sting studies. Fir.tly, we pr.vide a concise intr.duction to the r.sear.h backgr.und and outline the issue of steganogr.pher.identification. Secondly, we pr.sent fundamental concepts and techniques that establish a gener.l fr.mewor. for.identifying steganogr.pher.. Within this fr.mewor., state-of-the-ar. methods ar. summar.zed fr.m five key aspects: data acquisition, featur. ex.r.ction, featur. optimization, identification par.digm, and per.or.ance evaluation. Fur.her.or., theor.tical and ex.er.mental analyses ex.mine the advantages and limitations of these ex.sting methods. Finally, the sur.ey highlights outstanding issues in image steganogr.pher.identification that deser.e fur.her.r.sear.h. © 2024 The Author.. Published by Tech Science Pr.ss.

关键词： cover. communication Infor.ation hiding steganalysis steganogr.pher.identification steganogr.phy sur.ey