检索结果-内蒙古大学图书馆

第一届亚洲模式识别会议

作者： Zhong Zhang Wen Zhou Baihua Xiao Shuang Liu Chunheng Wang State Key Laboratory of Intelligent Control and Management of Complex Systems Institute of Automatio State Key Laboratory of Intelligent Control and Management of Complex SystemsInstitute of Automation

ISBN: (纸本)9781457701221

Recently, Independent Component Analysis based foreground detection has been proposed for indoor surveillance applications where the foreground tends to move slowly or remain still. Yet such a method often causes discrete segmented foreground objects. In this paper, we propose a novel foreground detection method named Contextual Constrained Independent Component Analysis (CCICA) to tackle this problem. In our method, the contextual constraints are explicitly added to the optimization objective function, which indicate the similarity relationship among neighboring pixels. In this way, the obtained de-mixing matrix can produce the complete foreground compared with the previous ICA model. In addition, our method performs robust to the indoor illumination changes and features a high processing speed. Two sets of image sequences involving room lights switching on/of and door opening/closing are tested. The experimental results clearly demonstrate an improvement over the basic ICA model and the image difference method.

关键词： foreground detection contextual constraints indoor surveillance independent component analysis

来源：评论

学校读者我要写书评

暂无评论

A Local Learning Based Image-To-Class Distance for Image Classification

A Local Learning Based Image-To-Class Distance for Image Cla...

引用

第一届亚洲模式识别会议

作者： Xinyuan Cai Rongguo Zhang Baihua Xiao Chunheng Wang State Key Laboratory of Intelligent Control and Management of Complex Systems Institute of Automatio State Key Laboratory of Intelligent Control and Management of Complex SystemsInstitute of Automation

ISBN: (纸本)9781457701221

Image-To-Class distance is first proposed in Naive- Bayes Nearest-Neighbor. NBNN is a feature-based image classifier, and can achieve impressive classification accuracy. However, the performance of NBNN relies heavily on the large number of training samples. If using small number of training samples, the performance will degrade. The goal of this paper is to address this issue. The main contribution of this paper is that we propose a robust Image-to-Class distance by local learning. We define the patch-to-class distance as the distance between the input patch to its nearest neighbor in one class, which is reconstructed in the local manifold space;and then our image-toclass distance is the sum of patch-to-class distance. Furthermore, we take advantage of large-margin metric learning framework to obtain a proper Mahalanobis metric for each class. We evaluate the proposed method on four benchmark datasets: Caltech, Corel, Scene13, and Graz. The results show that our defined Image-To-Class Distance is more robust than NBNN and Optimal-NBNN, and by combining with the learned metric for each class, our method can achieve significant improvement over previous reported results on these datasets.

关键词： local learning image-to-class distance large margin metric learning Na(i)ve Bayes Nearest-Neighbor image classification

来源：评论

学校读者我要写书评

暂无评论

Background Modeling by Exploring Multi-scale Fusion of Texture and Intensity in complex Scenes

Background Modeling by Exploring Multi-scale Fusion of Textu...

引用

第一届亚洲模式识别会议

作者： Zhong Zhang Wen Zhou Chunheng Wang Baihua Xiao Shuang Liu State Key Laboratory of Intelligent Control and Management of Complex Systems Institute of Automatio State Key Laboratory of Intelligent Control and Management of Complex SystemsInstitute of Automation

ISBN: (纸本)9781457701221

Background modeling is a fundamental yet challenging issue in video surveillance. Traditional methods usually adopt single feature type to solve the problem, while the performance is usually unsatisfactory when handling complex scenes. In this paper, we propose a multi-scale framework, which combines both texture and intensity feature, to achieve a robust and accurate solution. Our contributions are three folds: first, we provide a multi-scale analysis for the issue;second, for texture feature we propose a novel texture operator named Scale-invariant Centersymmetric Local Ternary Pattern, and a corresponding Pattern Adaptive Kernel Density Estimation technique for its probability estimation;third, we design a Simplified Gaussian Mixture Models for intensity feature. Our method is tested on several complex real world videos with illumination variation, soft shadows and dynamic backgrounds. The experimental results clearly demonstrate that our method is superior to the previous methods.

关键词： background modeling multi-scale fusion texture intensity

来源：评论

学校读者我要写书评

暂无评论

The application and issuse of linear wireless sensor networks

The application and issuse of linear wireless sensor network...

引用

2011 International Conference on System Science, Engineering Design and Manufacturing Informatization, ICSEM 2011

作者： Wang, Zhengjie Zhao, Xiaoguang Qian, Xu College of Information and Electrical Engineering Shandong University of Science and Technology Qingdao China State Key Laboratory of Intelligent Control and Management of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China Beijing China

ISBN: (纸本)9781457702471

Wireless sensor networks consist of a large number of sensor nodes that have low power and limited transmission range and can be used in various scenario. The nodes can be deployed in the long and narrow region, such as road, bridge, tunnel and pipeline, to get some interesting information. The linear topology of these network application is different other application and have special feature, such as multi-hop, long delay, long distance and low reliability. This paper introduces the concept of linear wireless sensor networks and discusses the classification of topology and key issue of this network. The application of the linear wireless sensor network, such as road, bridge, tunnel and pipeline is presented. The research challenges are discussed at last in this paper. © 2011 IEEE.

关键词： Topology

来源：评论

学校读者我要写书评

暂无评论

Analysis of neural interaction during adaptation of reach-to-grasp task under perturbation with Bayesian networks

Analysis of neural interaction during adaptation of reach-to...

引用

2011 4th International Conference on Biomedical Engineering and Informatics, BMEI 2011

作者： Sang, Dong Lv, Bin He, Huiguang Wang, Feiyue He, Jiping State Key Laboratory of Intelligent Control and Management of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China Harrington Departemnt of Bioengineering Center for Neural Interface Design Arizona State University Tempe United States

ISBN: (纸本)9781424493524

In this work, we took the analysis of neural interactions change in M1 of a monkey during the adaptation process for it to complete reach-to-grasp tasks with external perturbation across days. BN model was applied to model and evaluate neural interaction networks from recorded neural spike trains data of each set. Our results showed that for delay period across sets, interaction level of neural network tended to be higher during later stage of adaptation than during begin stage, which indicated the monkey performed more fully preparation through adaptation. In addition, for both delay period and peri-movement period, the neural interaction networks tended to change more stably from one set to the next as the monkey adapted to the perturbation experiment better. © 2011 IEEE.

关键词： Bayesian networks

来源：评论

学校读者我要写书评

暂无评论

Thrust analysis of the undulating ribbon-fin for biomimetic underwater robots

Thrust analysis of the undulating ribbon-fin for biomimetic ...

引用

International Conference on intelligent control and Information Processing

作者： Wang, Shuo Dong, Xiang Shang, Liu-Ji State Key Laboratory of Intelligent Control and Management of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing 100190 China Institute of Intelligent Machines Chinese Academy of Sciences P.O. Box 1130 Hefei City 230031 Anhui Province China

ISBN: (纸本)9781457708145

Biomimetic underwater robots have been paid more and more attention because of high efficiency, high maneuverability and low-noise. The undulating ribbon-fins used by rajiformes and gymnotiformes show better maneuverability than those of other fishes in turbulent water. So the undulating ribbon-fin propulsion is analyzed for biomimetic underwater robot design and optimization in the paper. The geometric model of the ribbon-fin is set up in Cartesian space. The undulating motion of the ribbon-fin is expressed in trigonometric functions. And the ribbon-fin has been divided into many small elements to compute the force acted on the small element. All the forces acted on the elements in one undulating period are summed, and the average thrust for one undulating period of the ribbon-fin is obtained. The simulation results show the influence of several important parameters on the ribbon-fin thrust. And an experimental platform is built and several experiments are carried out. And the difference between the results of simulation and experiment are discussed further. © 2011 IEEE.

关键词： Fins (heat exchange)

来源：评论

学校读者我要写书评

暂无评论

Implementation of a multi-link robotic dolphin with two 3-DOF flippers

引用

Journal of Computational Information systems 2011年第7期7卷 2601-2607页

作者： Shen, Fei Wei, Changming Cao, Zhiqiang Xu, De Yu, Junzhi Zhou, Chao State Key Laboratory of Intelligent Control and Management of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing 100190 China

A robotic dolphin with a pair of 3-DOF flippers, two turning units and a multi-link oscillatory propulsion mechanism is designed. The mechanical, hardware and software designs of the robotic dolphin are given. The flippers are fixed on both sides of the anterior body and each one is endowed with the ability of lead-lag, feathering, and up-down motions. The yawing of the robotic dolphin is realized via controlling the two turning units. The multi-link oscillatory propulsion mechanism consists of four joints, in which the first joint adopts the structure of worm gear and shaft driven by a DC motor in order to provide large torque and the other three oscillatory hinge joints are directly driven by servos. The propulsion is achieved via the coordinated controlling of the four joints. The performance of the robotic dolphin is well verified in the experiments. Copyright © 2011 Binary Information Press.

关键词： DC motors

来源：评论

学校读者我要写书评

暂无评论

Fraud-sensitive auction mechanism for sponsored search advertising

Fraud-sensitive auction mechanism for sponsored search adver...

引用

21st Workshop on Information Technologies and systems, WITS 2011

作者： Li, Linjing Zeng, Daniel Zhao, Huimin State Key Laboratory of Intelligent Control and Management of Complex Systems Institute of Automation Chinese Academy of Sciences China Department of Management Information Systems University of Arizona United States Sheldon B. Lubar School of Business University of Wisconsin Milwaukee United States

Generalized second-price (GSP) is currently the dominant auction mechanism used in the sponsored search advertising market. However, despite its tremendous commercial success and theoretical optimality, its effectiveness is jeopardized by the severe click frauds conducted by advertisers and third-party publishers and the vicious bidding strategy used by advertisers to exhaust the budget of rivals. In this paper, we analyze the drawbacks of GSP that tolerate or even encourage such negative behaviors (i.e., click fraud and vicious bidding) and propose a dynamic modification of the original GSP mechanism to address these drawbacks. Our modified auction mechanism incorporates budget into slot allocation and payment determination and relates the quality score of an advertisement to the current bid. Our analysis shows that our mechanism can effectively reduce the effects of click fraud and vicious bidding.

关键词： Marketing

来源：评论

学校读者我要写书评

暂无评论

A reinforcement learning based tag recommendation

引用

作者： Ge, Feng He, Yi Liu, Jin Lv, Xiaoming Zhang, Wensheng Li, Yiqun State Key Laboratory of Software Engineering Computer School Wuhan University 430072 China Logistics Group Wuhan University 430072 China State Key Laboratory of Intelligent Control and Management of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China Baidu Inc. 100085 China

ISBN: (纸本)9783642256578

This paper proposes a reinforcement learning based tag recommendation algorithm to deal with the data sparseness that affects the performance stability of collaborative filtering algorithms. Our algorithm integrates user tags into traditional collaborative filtering algorithms and attaching importance to user interest shift in the process of user interest learning process. Empirical Cases of comparing with traditional collaborative filtering algorithms indicate that our recommend algorithm exhibits better performance competition. © 2011 Springer-Verlag Berlin Heidelberg.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

A multi-target tracking approach combined with occlusion segmentation

A multi-target tracking approach combined with occlusion seg...

引用

International Workshop on advanced Computational Intelligence (IWACI)

作者： Huan Ding Wensheng Zhang State Key Laboratory of Intelligent Control and Management of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China

Multiple-target tracking in complex scenes is one of the most complicated problems in computer vision. Handling the occlusion between objects is the key issue in multiple target tracking. This paper presents an occlusion segmentation-based method to track multiple people in complex situations which are captured by static monocular cameras. In the proposed method, we calculate the probabilistic histogram of each object's optical flow vector, then use this motion statistic information along with the color and appearance information to construct a new expression of pixel distance. Finally, a stepwise classification and K-means clustering method are taken advantages of to accomplish occlusion segmentation. Object tracking is handled by a particle filter-based tracking framework, and a probabilistic appearance model is used to find the best particle. Experiments are conducted using public challenging data set PETS 2009. Results show that our approach can improve the performance of the existing tracking approach and handle dynamic occlusions better.

关键词： Computer vision Image motion analysis Probabilistic logic Optical reflection Target tracking Optical filters Histograms

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：