检索结果-内蒙古大学图书馆

Novel non-convex online algorithms for robust ranking

Journal of Information and Computational Science 2013年第17期10卷 5743-5759页

作者： Cheng, Fan Wang, Xufa Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Anhui University Hefei 230039 China School of Computer Anhui University Hefei 230039 China School of Computer University of Science and Technology of China Hefei 230027 China

Recently increasing attention has been focused on learning to rank, which aims to learn a ranking function from a set of training data with relevance lab.ls. Many of the ranking algorithms are based on the pairwise preference framework. However, one disadvantage of pairwise ranking is that it is sensitive to the outliers. To tackle the problem, in this paper, we propose two novel online algorithms based on non-convex Ramp Loss which can suppress the influence of outliers. The first one directly defines a new objective function based on Ramp Loss. Considering that the objective function is non-convex, we use the Concave-convex Procedure (CCCP) for convex approximation, then design an online algorithm based on Dual Coordinate Descend Method to learn the final ranking model. The second algorithm applies selective sampling technology in the training phase. It ignores the instances that lie in the flat region of the Ramp Loss in advance, before they are processed by the learner. Experimental results show that the algorithms we proposed have significant robustness to outliers and yield better generalization performance than the baseline rankers, and the results also prove that our algorithms have less computational running time for their less support vectors. © 2013 Binary Information Press.

关键词： Approximation algorithms

来源：评论

学校读者我要写书评

暂无评论

Image bag generator based on bag of visual words

引用

Journal of Information and Computational Science 2013年第5期10卷 1453-1462页

作者： Zhao, Shu Xu, Chao Xu, Xiansheng Xu, Chenchu Zhang, Yanping Ye, Hong School of Computer Science and Technology Anhui University Hefei 230039 China Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education Anhui University Hefei 230039 China Insurance Office Anhui Provincial Hospital Hefei 230001 China

Multi-instance learning is a good tool to deal with the ambiguity of image in Content-based Image Retrieval (CBIR). According to some researchers' work, image bag generator which transforms an image into an instance bag plays an important role in applying multi-instance learning for CBIR. In this paper, the bag generator based on bag of visual words, named Vdbag, is proposed. First, it extracts local features of image from all images, and then it obtains a set of visual words called visual vocabulary by clustering local features where each local cluster's center is a visual word. Our work mainly includes (1) We introduce bag of visual words into image bag generator. (2) We give a reasonable description to express the instance in an image bag. We add the frequency of each visual word as the last dimension of this visual word to construct the corresponding instance. Experimental results with Corel image database show that the proposed method could achieve better precision than that of some existing bag generators. © 2013 Binary Information Press.

关键词： Content based retrieval

来源：评论

学校读者我要写书评

暂无评论

Method for convergent projective non-negative matrix factorization

引用

Journal of Information and Computational Science 2013年第6期10卷 1847-1854页

作者： Hu, Lirui Wu, Jianguo Wang, Lei Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education Anhui University Hefei 230039 China School of Computer Science and Technology Nantong University Nantong 226019 China School of Computer Science and Technology Anhui University Hefei 230039 China

In order to solve the problem of algorithm convergence in Projective Non-negative Matrix Factorization (P-NMF), a method, called Convergent Projective Non-negative Matrix Factorization (CP-NMF), is proposed. In CP-NMF, an objective function of Frobenius norm is defined. The Taylor series expansion and the Newton iteration formula of solving root are used. An iterative algorithm for basis matrix is derived, and a proof of algorithm convergence is provided. Experimental results show that the convergence speed of the algorithm is higher, however it is affected by the initial value of the basis matrix;relative to Non-negative Matrix Factorization (NMF), the orthogonality and the sparseness of the basis matrix are better, however the reconstructed results of data show that the basis matrix is still approximately orthogonal;in face recognition, there is higher recognition accuracy. The method for CP-NMF is effective. Copyright © 2013 Binary Information Press.

关键词： Face recognition

来源：评论

学校读者我要写书评

暂无评论

Application and method of feature extraction for approximate orthogonal non-negative matrix factorization

引用

Journal of Computational Information Systems 2013年第2期9卷 805-812页

作者： Hu, Lirui Wu, Jianguo Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education Anhui University Hefei 230039 China School of Computer Science and Technology Nantong University Nantong 226019 China School of Computer Science and Technology Anhui University Hefei 230039 China

In order to get effectively feature vectors of samples, a method of feature extraction is proposed based on non-negative matrix factorization (NMF) with approximate orthogonal constraint. An objective function is defined to impose approximate orthogonal constraint, in addition to the non-negative constraint in the standard NMF. An algorithm of iterative update for basis matrix and weight matrix in non-negative matrix factorization is derived, and then computational method of feature vectors is gotten. A proof of the convergence of the algorithm is provided. Empirical results of handwritten character recognition show that there is high recognition accuracy if only the rank of the basis matrix is set correctly. The method of feature extraction is useful to get essential feature of sample, and is effective to recognition when the structures of the handwritten characters are obviously different. Copyright © 2013 Binary Information Press.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Multi-granular model to solve netted problem

引用

Journal of Computational Information Systems 2013年第19期9卷 7919-7927页

作者： Chen, Jie Zhang, Yanping Zhao, Shu Liu, Lulu School of Computer Science and Technology Anhui University Hefei 230039 China Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education Anhui University Hefei 230039 China Department of Computer Science and Technology Hefei Normal University Hefei 230061 China

The complex problems usually have netted structure, namely netted problems. The generally solving methods are based on sequence-searching or tree-searching of this kind of problems. But the structure of netted problems isn't merely sequence or tree. The Multi-granular Model was proposed in this paper which is put forward based on granular transaction of Quotient Space Theory. The whole problem is represented as a directed graph. Firstly, this graph was projected by constraints of original problem into coarse-grained spaces. Then granular decomposition is used to clear up the circle paths in new graph. In optimal quotient space, we search the answers according to the ordered sequence which is got by traversing the result of granular decomposition. In this paper, an applicable example is given and experiment contrasted with traditional tree searching method shows that this algorithm spends less time complexity and smaller space complexity. © 2013 Binary Information Press.

关键词： Directed graphs

来源：评论

学校读者我要写书评

暂无评论

A quasi-static clustering algorithm in mobile heterogeneous sensor network

引用

Journal of Computational Information Systems 2013年第9期9卷 3741-3749页

作者： Wang, Bingting Lin, Qibin Hu, Yanjun Liang, Hui Wang, Xianbing Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Anhui University Hefei 230039 China School of Mechanical and Electronic Engineering Chuzhou University Chuzhou 239000 China School of Information Science and Technology Southwest Jiaotong University Chengdu 610031 China

In this paper, we constructed the architecture of mobile heterogeneous sensor network by introducing user equipments (UEs) in static wireless sensor network. We analyzed the failure of traditional static clustering algorithm, gave a set of parameters that affected the performance of cluster heads in the mobile environment and proposed a Quasi-Static Clustering Algorithm (QSCA). Through theoretical analysis and simulation, this algorithm can reduce average delay and prolong network lifetime. We verified the adaptability of QSCA algorithm to energy consumption and dynamic topology by comparing with the LEACH algorithm. © 2013 by Binary Information Press.

关键词： Clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

Wavelet denoising of SAR images corrupted by speckle noise using cycle spinning

Wavelet denoising of SAR images corrupted by speckle noise u...

引用

2nd International Conference on intelligent Materials, Applied Mechanics and Design Science, IMAMD 2013

作者： Ji, Ming Wei Liu, Yan Li Zhang, De Xiang The Institute of Mechanical and Electrical Engineering of Anhui University of Architecture Hefei 230601 China Key Lab. of Intelligent Computing and Signal Processing Anhui University Hefei 230039 China

ISBN: (纸本)9783037857175

A novel and efficient speckle noise reduction algorithm based on wavelet transform by cycle spinning for removing speckle of unknown variance and minimizing the effect of pseudo-Gibbs phenomena from Synthetic Aperture Radar (SAR) images is proposed. Therefore, we show that the sub-band decompositions of logarithmically transformed SAR images. Then, we process and reconstruct multi-resolution wavelet coefficients by wavelet-threshold using cycle spinning, a technique estimating the true images as the linear average of individual estimates derived from wavelet thresholded translated versions of the noise images. Experimental results show that the proposed de-noising algorithm is possible to achieve an excellent balance between suppresses speckle effectively and weaken as many image Gibbs phenomena as possible. Quantitative and qualitative comparisons of the results obtained by the new method with the results achieved from the other speckle noise reduction techniques demonstrate its higher performance for speckle reduction in SAR images. © (2013) Trans Tech Publications, Switzerland.

关键词： Wavelet transforms

来源：评论

学校读者我要写书评

暂无评论

Uniplanar coupled-feed printed PIFA for LTE/WWAN operation in the laptop computer

Uniplanar coupled-feed printed PIFA for LTE/WWAN operation i...

引用

2013 International Conference on Communication Technology (ICCT 2013)

作者： Sai Zhang Minquan Li Bo Rong Haixing Chang Meng Peng You Ding Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Anhui University

A new uniplanar printed PIFA with a coupled-feed structure for LTE/WWAN operation is proposed. The PIFA consists of a coupled-feed loop structure and two separated radiators which are able to generate additional bands for the operation. The loop antenna path and the longer radiator can produce a dualresonance generating the required lower band exactly, while three wide bands contributed by the loop path, the loop path with the longer radiator and the loop path with the shorter radiator can form an operating band to cover the(1710–2690 MHz) band. Because of the small size of only 14×70 mm2, the antenna which can cover the required eight bands is therefore very suitable for the modern Ultra-thin laptops.

关键词： PIFA coupled-feed LTE/WWAN operation Ultra-thin laptops

来源：评论

学校读者我要写书评

暂无评论

Video summarization by robust low-rank subspace segmentation

Video summarization by robust low-rank subspace segmentation

引用

8th International Conference on Bio-Inspired computing: Theories and Applications, BIC-TA 2013

作者： Tu, Zhengzheng Sun, Dengdi Luo, Bin Key Lab of Intelligent Computing and Signal Processing of Ministry of Education Anhui University Hefei230039 China School of Computer Science and Technology Anhui University Hefei230601 China Key Lab of Industrial Image Processing and Analysis of Anhui Province Hefei230039 China

ISBN: (纸本)9783642375019

Video summarization provides condensed and succinct representations of the content of a video stream. A static storyboard summarization approach based on robust low-rank subspace segmentation is proposed in this paper. Firstly, video frames are represented as multi-dimensional vectors, and then embedded into a group of affine subspaces using low-rank representation according to the content similarity of the frames in the same subspace. Secondly, a series of subspaces are segmented based on the Normalized Cuts algorithm. The video summary is finally generated by choosing key frames from the significant subspaces and ranking these key frames in temporal order. The experimental results demonstrate that the proposed summarization algorithm can produce crucial key frames and effectively reduce the visual content redundancy in summary comparing with the conventional approaches. © Springer-Verlag Berlin Heidelberg 2013.

关键词： Vectors

来源：评论

学校读者我要写书评

暂无评论

Automatic image annotation using semantic relevance 13

Automatic image annotation using semantic relevance

引用

5th International Conference on Internet Multimedia computing and Service, ICIMCS 2013

作者： Zhao, Peng Lu, Yijuan Wang, WenBin Zhu, WeiWei Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Anhui University Hefei 230601 China Dept. of Computer Science Texas State University San Marcos TX 78666 United States

ISBN: (纸本)9781450322522

Due to the semantic gap between low-level visual feature and high-level semantic concept, image annotation plays an important role in image retrieval. In this paper, an automatic image annotation approach using semantic relevance is proposed. It constructs an improved probabilistic model to characterize different regions' contributions to the semantics more accurately based on the spatial, visual and contextual information of the region. And it also helps expand the coverage of the semantic concept with semantic relevance information. The performance of the proposed approach has been evaluated on the standard Corel dataset. The experimental results have demonstrated its potential and effectiveness. © 2013 ACM.

关键词： Image annotation

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：