检索结果-南通市图书馆

Learning Domain-Invariant and Discriminative Features for Homogeneous Unsupervised Domain Adaptation

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2020年第6期29卷 1119-1125页

作者： ZHANG yun WANG Nianbin cai Shaobin College of Computer Science and Technology Harbin Engineering University School of Electronic Information engineering Foshan University

A classifier trained on the label-rich source dataset tends to perform poorly on the unlabeled target dataset because of the distribution discrepancy across different *** domain adaptation aims to transfer knowledge from the labeled source dataset to the unlabeled target dataset to solve this *** of the existing unsupervised domain adaptation methods only concentrate on learning domain-invariant features across different domains,but they neglect the discriminability of the learned features to satisfy the cluster *** this paper,we propose Semantic pairwise centroid alignment(SPCA),which is a point-wise method to learn both domain-invariant and discriminative features for homogeneous unsupervised domain *** utilizes a novel semantic centroid loss to reduce the intraclass distance in feature space by using source data and target High-confidence centroid points(HCCPs).Then a classifier trained on source features is expected to generalize well on target *** experiments on visual recognition tasks verify the effectiveness of the proposed SPCA and also demonstrate that both domaininvariant and discriminative features learned by SPCA can significantly boost the performance of homogeneous unsupervised domain adaptation.

关键词： HCCP learning domain-invariant features Knowledge engineering techniques visual recognition SPCA Image recognition point-wise method semantic pairwise centroid alignment unlabeled target dataset label-rich source dataset unsupervised learning high-confidence centroid points Computer vision and image processing techniques image classification domaininvariant features homogeneous unsupervised domain adaptation methods

Value Iteration-Based Cooperative Adaptive Optimal Control for Multi-Player Differential Games With Incomplete Information

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

IEEE/CAA Journal of Automatica Sinica 2024年第3期11卷 690-697页

作者： yun Zhang Lulu Zhang yunze cai the Department of Automation Shanghai Jiao Tong UniversityShanghai 200240 the Key Laboratory of System Control and Information Processing Ministry of Education of ChinaShanghai 200240China

This paper presents a novel cooperative value iteration(VI)-based adaptive dynamic programming method for multi-player differential game models with a convergence *** players are divided into two groups in the learning process and adapt their policies *** method removes the dependence of admissible initial policies,which is one of the main drawbacks of the PI-based ***,this algorithm enables the players to adapt their control policies without full knowledge of others’ system parameters or control *** efficacy of our method is illustrated by three examples.

关键词： Adaptive dynamic programming incomplete information multi-player differential game value iteration

Compressed data separation via dual frames based split-analysis with Weibull matrices

在线全文

学校读者我要写书评

暂无评论

Applied Mathematics(A Journal of Chinese Universities) 2013年第4期28卷 427-437页

作者： cai yun LI Song Department of Mathematics Zhejiang University

In this paper, we consider data separation problem, where the original signal is composed of two distinct subcomponents, via dual frames based Split-analysis approach. We show that the two distinct subcomponents, which are sparse in two diff erent general frames respectively, can be exactly recovered with high probability, when the measurement matrix is a Weibull random matrix （not Gaussian） and the two frames satisfy a mutual coherence property. Our result may be significant for analysing Split-analysis model for data separation.

关键词： Compressed sensing data separation dual frames split-analysis Weibull random matrices

Convergence analysis of projected gradient descent for Schatten-p nonconvex matrix recovery

在线全文

学校读者我要写书评

暂无评论

Science China Mathematics 2015年第4期58卷 845-858页

作者： cai yun LI Song Department of Mathematics Zhejiang University

The matrix rank minimization problem arises in many engineering applications. As this problem is NP-hard, a nonconvex relaxation of matrix rank minimization, called the Schatten-p quasi-norm minimization(0 en developed to approximate the rank function closely. We study the performance of projected gradient descent algorithm for solving the Schatten-p quasi-norm minimization(0 ence guarantee and error bound for this algorithm and show that the algorithm is robust to noise with an exponential convergence rate.

关键词： low rank matrix recovery nonconvex matrix recovery projected gradient descent restricted isometry property

New multi-camera calibration algorithm based on 1D objects

在线全文

学校读者我要写书评

暂无评论

Journal of Zhejiang University-Science A(Applied Physics & engineering) 2008年第6期9卷 799-806页

作者： Zi-jian ZHAO yun-cai LIU Institute of linage Processing and Pattern Recognition Shanghai Jiao Tong University Shanghai 200240 China

A new calibration algorithm for multi-camera systems using 1D calibration objects is proposed. The algorithm inte- grates the rank-4 factorization with Zhang (2004)'s method. The intrinsic parameters as well as the extrinsic parameters are re- covered by capturing with cameras the 1D object's rotations around a fixed point. The algorithm is based on factorization of the scaled measurement matrix, the projective depth of which is estimated in an analytical equation instead of a recursive form. For more than three points on a 1D object, the approach of our algorithm is to extend the scaled measurement matrix. The obtained parameters are finally refined through the maximum likelihood inference. Simulations and experiments with real images verify that the proposed technique achieves a good trade-off between the intrinsic and extrinsic camera parameters.

关键词： Multi-camera calibration Homography Factorization Scaled measurement matrix Projective depth

The influences of convective overshooting and semiconvection on the chemical evolution of massive stars

在线全文

学校读者我要写书评

暂无评论

Research in Astronomy and Astrophysics 2014年第8期14卷 979-991页

作者： cai-yun Ding Yan Li yunnan Observatories Chinese Academy of Sciences Key Laboratory for the Structure and Evolution of Celestial Objects Chinese Academy of Sciences University of Chinese Academy of Sciences

In massive stars, convection in the interior is different from that of inter- mediate and small mass stars. In the main-sequence phase of small mass stars, there is a convective core and a radiative envelope, between which are the radiative inter- mediate layers with uneven chemical abundances. Semiconvection would occur in the intermediate layers between the convective core and the homogeneous envelope in massive stars. We treat core convective overshooting and semiconvection together as a process. We found that when decreasing overshooting, the semiconvection is more pronounced. In these two processes, we introduce one diffusive parameter D, which is different from other authors who have introduced different parameters for these two zones. The influences of the turbulent diffusion process on chemical evolution and other quantities of the stellar structure are shown in the present paper.

关键词： stars evolution -- stars abundances -- convection -- diffusion

Visualization of COVID-19 spread based on spread and extinction indexes

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2020年第6期63卷 220-222页

作者： Song-Hai ZHANG yun cai Jian LI Department of Computer Science and Technology Tsinghua University Beijing National Research Center for Information Science and Technology(BNRist)

From December 2019, Coronavirus disease 2019(COVID-19) rapidly spread from the city of Wuhan (Hubei Province) to all of China, which was exacerbated by mass population movements associated with the Chinese Spring Festival. To date, more than 80000 people have been infected,and more than 3000 have died. In addition,COVID-19 has already spread to 142 countries,

关键词： Visualization of COVID-19 spread based on spread and extinction indexes

Equilibrium of Organic Matter in Heavy Fraction for Three Long-term Experimental Field Soils in China

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Pedosphere 2006年第2期16卷 177-184页

作者： YIN yun-Feng cai Zu-Cong State Key Laboratory of Soil and Sustainable Agriculture Institute of Soil Science Chinese Academy of Sciences Nanjing 210008 （China）.

Considerable evidence that the soil organic matter （OM） level in agricultural soils will gradually over time reach an equilibrium state under certain bioclimatic conditions and for certain cropping systems has been accumulating. Although models or long-term experiments have been used, this research used physical fractionation procedure to attain an soil OM equilibrium value. To obtain soil OM equilibrium values in the heavy fraction, typical soils from three long-term field experiments at Fengqiu and Yingtan State Key Agro-Ecological Experimental Stations in China were studied using a simple density fractionation procedure and employing the Langmuir equation. Results for the fluvo-aquic soil with organic fertilizer treatments indicated that the soil OM equilibrium value in the heavy fraction was twofold more than that in the inorganic treatments; however, for the paddy soil developed on red soil the OM equilibrium value in the heavy fraction for both treatments was almost identical. It suggested that for fiuvo-aquic soils the increased potential of OM for the heavy fraction in the long run was larger for the organic than the inorganic fertilizer applications, whereas for paddy soils developed on red soils under the same conditions the present OM content in the heavy fraction was at or close to this equilibrium level for all treatments, and increased potential was very limited.

关键词： equilibrium value heavy fraction Langmuir equation soil organic matter

Characteristic Modules of Dual Extensions and Grobner Bases

在线全文

学校读者我要写书评

暂无评论

Acta Mathematica Sinica,english Series 2004年第6期20卷 1119-1130页

作者： yun Ge XU Long cai LI Faculty of Mathematics and Computer Science Hubei University Academy of Mathematics and Systems Science Chinese Academy of Sciences

Let C be a finite dimensional directed algebra over an algebraically closed field k and A=A(C) the dual extension of *** characteristic modules of A are constructed explicitly for a class of directed algebras,which generalizes the results of ***,it is shown that the characteristic modules of dual extensions of a certain class of directed algebras admit the left Gr(?)bner basis theory in the sense of ***.

关键词： Quasi-hereditary algebra Dual extension Characteristic module (Left)Gr(?)bner basis