检索结果-内蒙古大学图书馆

arXiv 2017年

作者： Yu, Jun Shi, Shengjie Gao, Fei Tao, Dacheng Huang, Qingming Key Laboratory of Complex Systems Modeling and Simulation School of Computer Science and Technology Hangzhou Dianzi University Hangzhou310018 China State Key Laboratory of Integrated Services Networks Xidian University Xian710071 China UBTech Sydney Artificial Intelligence Institute The School of Information Technologies Faculty of Engineering and Information Technologies The University of Sydney J12 Cleveland St DarlingtonNSW2008 Australia School of Computer and Control Engineering University of Chinese Academy of Sciences Beijing100190 China

Face photo-sketch synthesis aims at generating a facial sketch/photo conditioned on a given photo/sketch. It is of wide applications including digital entertainment and law enforcement. Precisely depicting face photos/sketches remains challenging due to the restrictions on structural realism and textural consistency. While existing methods achieve compelling results, they mostly yield blurred effects and great deformation over various facial components, leading to the unrealistic feeling of synthesized images. To tackle this challenge, in this work, we propose to use the facial composition information to help the synthesis of face sketch/photo. Specially, we propose a novel composition-aided generative adversarial network (CA-GAN) for face photo-sketch synthesis. In CA-GAN, we utilize paired inputs including a face photo/sketch and the corresponding pixel-wise face labels for generating a sketch/photo. In addition, to focus training on hard-generated components and delicate facial structures, we propose a compositional reconstruction loss. Finally, we use stacked CA-GANs (SCA-GAN) to further rectify defects and add compelling details. Experimental results show that our method is capable of generating both visually comfortable and identity-preserving face sketches/photos over a wide range of challenging data. Our method achieves the state-of-the-art quality, reducing best previous Fréchet Inception distance (FID) by a large margin. Besides, we demonstrate that the proposed method is of considerable generalization ability. We have made our code and results publicly available: Https://***/ca-gan/. Copyright © 2017, The Authors. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

A novel genetic algorithm for constructing uniform test forms of cognitive diagnostic models

A novel genetic algorithm for constructing uniform test form...

引用

Congress on Evolutionary Computation

作者： Ye-shi Jiang Ying Lin Jing-Jing Li Zheng-jia Dai Jun Zhang Xinglin Zhang Ministry of Education Key Laboratory of Machine Intelligence - & Sensor Networks P.R. China School of Data & Computer Science Sun Yat-sen University School of Computer Science South China Normal University Department of Psychology Sun Yat-sen University State Key Laboratory of Mathematical Engineering and Advanced Computing School of Comouter Science & Engineering South China University of Technology

Cognitive diagnostic models (CDMs) are a new class of test models developed for educational assessment. They have gained growing attention in recent years for their distinctive ability to provide detailed feedback about examinees' ability. Automatic test assembly (ATA), as in other test models, has been one of the most critical issues in the development and applications of CDMs. However, developing ATA methods for CDMs is especially challenging because no close-form expressions can measure the quality of a test form based on the items used. Although some heuristic methods have been proposed for building a single test form of CDMs, few ATA methods can construct uniform test forms of CDMs, in which each test form contains a different set of items but meets equivalent demand of test quality. In order to fill the gap, this paper proposes a novel genetic algorithm (GA) for constructing uniform test forms of CDMs. The effectiveness and efficiency of the proposed method is validated on a synthetic item pool under different conditions.

关键词： Genetic algorithms Error analysis Biological cells Computational modeling Sociology Statistics computer science

来源：评论

学校读者我要写书评

暂无评论

Dynamics of Interacting Diseases

引用

Physical Review X 2014年第4期4卷 041005-041005页

作者： Joaquín Sanz Cheng-Yi Xia Sandro Meloni Yamir Moreno Institute for Biocomputation and Physics of Complex Systems (BIFI) University of Zaragoza Zaragoza 50018 Spain Department of Theoretical Physics University of Zaragoza Zaragoza 50009 Spain Key Laboratory of Computer Vision and System (Ministry of Education) and Tianjin Key Laboratory of Intelligence Computing and Novel Software Technology Tianjin University of Technology Tianjin 300191 People’s Republic of China Complex Networks and Systems Lagrange Lab Institute for Scientific Interchange Turin 10126 Italy

Current modeling of infectious diseases allows for the study of complex and realistic scenarios that go from the population to the individual level of description. However, most epidemic models assume that the spreading process takes place on a single level (be it a single population, a metapopulation system, or a network of contacts). In particular, interdependent contagion phenomena can be addressed only if we go beyond the scheme-one pathogen-one network. In this paper, we propose a framework that allows us to describe the spreading dynamics of two concurrent diseases. Specifically, we characterize analytically the epidemic thresholds of the two diseases for different scenarios and compute the temporal evolution characterizing the unfolding dynamics. Results show that there are regions of the parameter space in which the onset of a disease’s outbreak is conditioned to the prevalence levels of the other disease. Moreover, we show, for the susceptible-infected-susceptible scheme, that under certain circumstances, finite and not vanishing epidemic thresholds are found even at the limit for scale-free networks. For the susceptible-infected-removed scenario, the phenomenology is richer and additional interdependencies show up. We also find that the secondary thresholds for the susceptible-infected-susceptible and susceptible-infected-removed models are different, which results directly from the interaction between both diseases. Our work thus solves an important problem and paves the way toward a more comprehensive description of the dynamics of interacting diseases.

关键词： Complex systems Interdisciplinary physics Statistical physics

来源：评论

学校读者我要写书评

暂无评论

A set-based locally informed discrete particle swarm optimization

A set-based locally informed discrete particle swarm optimiz...

引用

15th Annual Conference on Genetic and Evolutionary Computation, GECCO 2013

作者： Ma, Yun-Yang Gong, Yue-Jiao Chen, Wei-Neng Zhang, Jun Department of Computer Science Key Laboratory of Machine Intelligence and Sensor Networks Sun Yat-sen University China

ISBN: (纸本)9781450319645

This paper proposed an efficient discrete PSO algorithm. Following the general process of the recently proposed locally informed particle swarm (LIPS), the velocity update of each particle in the proposed algorithm depends on the pbests of its nearest neighbors. However, in order to achieve optimization in discrete space, the related arithmetic operators and the concept of 'distance' in LIPS are redefined based on set theory. Thus, the proposed algorithm is termed Set-based LIPS (S-LIPS). Moreover, a reset scheme is embedded in S-LIPS to further improve population diversity in S-LIPS. By using the locally informed update mechanism and the reset scheme, the proposed algorithm is able to have both high convergence speed and good global search ability. S-LIPS is compared with a set-based comprehensive learning PSO on TSP benchmark instances. The experimental result shows that S-LIPS is a very promising algorithm for solving discrete problems, especially in the case where the scale of the problem is large.

关键词： Traveling salesman problem

来源：评论

学校读者我要写书评

暂无评论

Space-based initialization strategy for particle swarm optimization

Space-based initialization strategy for particle swarm optim...

引用

15th Annual Conference on Genetic and Evolutionary Computation, GECCO 2013

作者： Yin, Liang Hu, Xiao-Min Zhang, Jun Department of Computer Science Key Laboratory of Machine Intelligence and Sensor Networks Sun Yat-sen University China

ISBN: (纸本)9781450319645

Particle Swarm Optimization (PSO) is a population-based stochastic optimization algorithm that has been applied to various scientific and engineering problems. Despite its fast convergence speed, the original PSO is easy to fall into local optima when solving multi-modal functions. To address this problem, we present a novel initialization strategy, namely Space-based Initialization Strategy (SIS), to help PSO avoid local optima. We embed SIS into the standard PSO and form a novel PSO variant named SIS-PSO. The performance of SIS-PSO is validated by 13 benchmark functions and the experimental results demonstrate that the SIS enables PSO to achieve faster convergence speed and higher solution accuracy especially in multi-modal problems.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

A novel pheromone-based evolutionary algorithm for solving degree-constrained minimum spanning tree problem

A novel pheromone-based evolutionary algorithm for solving d...

引用

15th Annual Conference on Genetic and Evolutionary Computation, GECCO 2013

作者： Huang, Xiao-Ma Gong, Yue-Jiao Zhang, Jun Department of Computer Science Key Laboratory of Machine Intelligence and Sensor Networks Sun Yat-sen University China

ISBN: (纸本)9781450319645

The degree-constrained minimum spanning tree problem (dc- MSTP) is crucial in the design of networks and it is proved to be NP-hard. The recently developed evolutionary algorithm utilizing node-depth-degree representation (EANDD) has successfully enabled the dc-MSTP solvable by generating new spanning trees in average time complexity O(√ n) , which is the fastest in the literature. However, as the generic operation of EANDD is to change two edges that are randomly selected from the entire tree, the efficiency of EANDD still has potential to be further improved. In this paper, we propose a novel pheromone-based tree modification method (PTMM) to improve the efficiency of EANDD. For each edge, a pheromone value is defined based on the historical contribution of the edge to the fitness of the spanning tree. Then, PTMM considers the pheromone value on each edge as a desirability measure for selecting the edge to construct the spanning tree. In this way, the more promising edge is more likely to be selected and therefore the efficiency of the tree modification operation in EANDD can be improved. The effectiveness and effieciency of PTMM is demonstrated on a set of benchmark instances in comparison with the original EANDD.

关键词： Evolutionary algorithms

来源：评论

学校读者我要写书评

暂无评论

Efficient Information Dissemination in Dynamic networks

Efficient Information Dissemination in Dynamic Networks

引用

42nd Annual International Conference on Parallel Processing (ICPP)

作者： Yang, Zhiwei Wu, Weigang Chen, Yishun Zhang, Jun Dept. of Computer Science Sun Yat-sen University Guangzhou 510006 China College of Computational Science Zhongkai University of Agriculture and Engineering Guangzhou 510006 China Ministry of Education Key Laboratory of Machine Intelligence and Sensor Networks Guangzhou 510006 China

ISBN: (纸本)9780769551173

Dynamic network is the abstraction of networks with frequent topology changes arising from node mobility or other reasons. With the model of dynamic network, distributed computation problems can be formally studied with rigorous correctness. Information dissemination is one of such problems and it has received much attention recently. However, existing works focus on the time cost of dissemination, i.e. how fast the information can be disseminated to all nodes, and communication cost has been largely ignored. Our work focuses on high communication efficiency in information dissemination with correctness property guaranteed. We achieve this by making use of cluster-based hierarchy. Clustering has been widely studied and used in wireless networks to reduce communication cost. However, to the best of our knowledge, it has never been considered in the study of dynamic networks. In this paper, we firstly propose a dynamic network model, named (T, L)-HiNet, to extend existing dynamic network model with clusters. (T, L)-HiNet includes several properties defining the dynamics of cluster hierarchy in a dynamic network. Base on (T, L)-HiNet, we design hierarchical information dissemination algorithms for different scenarios of dynamics. The correctness of our algorithms is proved and the performance is analyzed. Compared with the algorithm recently proposed by Kuhn, Lynch and Oshman [7], our design can significantly reduce communication and our objective is fully achieved.

关键词： Information dissemination dynamic network cluster distributed algorithm system model

来源：评论

学校读者我要写书评

暂无评论

Robust and Dynamic Data Aggregation in Wireless Sensor networks: A Cross-Layer Approach

Robust and Dynamic Data Aggregation in Wireless Sensor Netwo...

引用

Symposia and Workshops on Ubiquitous, Autonomic and Trusted Computing, UIC-ATC

作者： Weigang Wu Jiannong Cao Hejun Wu Jingjing Li The Key Laboratory of Machine Intelligence and Sensor Networks Ministry of Education China Department of Computer Science Sun Yat-sen University Guangzhou China Department of Computing Hong Kong Polytechnic University Hong Kong China

In-network data aggregation is an effective technique to reduce communication cost in wireless sensor networks. Recent works have focused on two issues individually: dynamic aggregation to handle irregular traffic of events and robust aggregation to tolerate packet losses. However, how to achieve both the objectives simultaneously is still not touched. In this paper, we propose a cross-layer approach to robust and dynamic data aggregation by making use of direct support from MAC layer. A new MAC protocol, DA-MAC is delicately designed to serve such purpose. Based on the channel contention information obtained from DA-MAC, a node can dynamically determine where and when to do aggregation. To cope with packet losses, a virtual overlay, Rings is adopted to forward one packet to multiple nodes. We have implemented our design in TinyOS based sensor networks. Performance evaluations though simulations and experiments show that, compared with existing algorithms, our proposed solution is more efficient in terms of both time and energy cost.

关键词： Heuristic algorithms Media Access Protocol Robustness Wireless sensor networks Algorithm design and analysis

来源：评论

学校读者我要写书评

暂无评论

Stationary properties of random direction models

Stationary properties of random direction models

引用

2nd International Conference on networks Security, Wireless Communications and Trusted Computing, NSWCTC 2010

作者： Liu, Yan-Tao Key Laboratory of Computer Networks and Intelligence College of Information Science and Engineering Bohai University Jinzhou China

ISBN: (纸本)9780769540115

Stationary properties of node distribution and speed distribution for random direction model were studied. Firstly, geometric probability was applied to analyze stationary node distribution. A closed form probability density function for circular regions was obtained which indicated that mobile nodes tended to disperse with simulation advancing. Secondly, we applied palm calculus to calculate probability density function of stationary speed distribution and proved that there was speed decay phenomenon in this model. Additionally, a general explanation was given to such issue. © 2010 IEEE.

关键词： Probability density function

来源：评论

学校读者我要写书评

暂无评论

Analysis to random direction model of ad-hoc networks

Analysis to random direction model of ad-hoc networks

引用

作者： Liu, Yan-Tao Lu, Ji-Hua Liu, Heng Key Laboratory of Computer Networks and Intelligence College of Information Science and Engineering Bohai University Liaoning Province China Department of Electrical Engineering School of Information Science and Electronics Beijing Institute of Technology Beijing China

The asymptotic properties of node distribution and speed distribution in random direction model were analyzed, respectively, by the tools of geometric probability and palm calculus. The probability density function for node distribution in circular regions was obtained which indicated that mobile nodes tended to disperse as simulation advancing. The speed decay phenomenon was confirmed in this model. Moreover, the hypostasis of speed decay was proved to be the correlation between speed and duration within any movement period. Copyright © 2010 The Institute of Electronics, Information and Communication Engineers.

关键词： Calculations

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：