检索结果-内蒙古大学图书馆

Speech separation in the vehicle environment based on fastICA algorithm

Journal of Multimedia 2012年第1期7卷 33-40页

作者： Zhang, Jindong Qin, Guihe Liu, Ye College of computer science and technology Jilin University Changchun China Key Laboratory of Symbol Computation Knowledge Engineering of the Ministry of Education Changchun China

The speech interaction in-vehicle was mainly realized by the speech recognition. The human-machine interaction around was usually disturbed by the noise, and the speech received by the receiver was not the original pure speech, so compared to the pure environment, the accuracy of the speech recognition declined so sharply that it could not meet the demand of the practical application of human- machine interaction. So the speech recognition was required to have the strong adaptability and processing capacity to the speech with noise. In this paper, the FastICA algorithm in signal process and statistics was studied and used to separate the driver's speech in the vehicle environment, and realizes the pretreatment in recognizing driver's speech. The effectiveness of the method had been validated in actual vehicle experimental configuration. © 2012 Academy Publisher.

关键词： Speech processing

来源：评论

学校读者我要写书评

暂无评论

Improving the Recommendation of Collaborative Filtering by Fusing Trust Network

Improving the Recommendation of Collaborative Filtering by F...

引用

International Conference on computational Intelligence and Security

作者： Bo Yang Pengfei Zhao Shuqiu Ping Jing Huang School of Computer Science and Technology Key Laboratory of Symbol Computation and Knowledge Engineering of the Ministry of Education Jilin University Changchun P.R.China

To accurately and actively provide users with their potentially interested information or services is the main task of a recommender system. Collaborative filtering is one of the most widely adopted recommender methods, whereas it is suffering the issue of sparse rating data that will severely degenerate the quality of recommendations. To address this issue, the article proposes a novel method, named the FTRA (Fusing Trust and Ratings), trying to improve the performance of collaborative filtering recommendation by means of elaborately integrating twofold sparse information, i.e., the conventional rating data given by users and the social trust network among the same users. The performance of FTRA is rigorously validated by comparing it with six representative methods on a real-world dataset. The experimental results show that the FTRA outperforms all other competitors in terms of both precision and recall. More importantly, our work suggests that the strategy of augmenting sparse rating data by fusing trust networks does significantly improve the quality of conventional collaborative filtering recommendation, and its quality could be further improved by means of designing more effective integrating schemes.

关键词： Collaboration Recommender systems Measurement Social network services Web sites Educational institutions

来源：评论

学校读者我要写书评

暂无评论

A hybrid optimization algorithm for the multi-chromosomal reversal median problem

A hybrid optimization algorithm for the multi-chromosomal re...

引用

1st International Conference on engineering and Technology Innovation, ICETI 2011

作者： Zhang, Yi Zhang, Meng Department of Computer Science Jilin Business and Technology College Changchun 130062 China Military Simulation Technology Institute Air Force Aviation University Changchun130022 China Key Laboratory of Symbol Computation and Knowledge Engineering of Ministry of Education College of Computer Science and Technology Jilin University Changchun 130012 China

ISBN: (纸本)9783037853245

In this paper, we introduce a hybrid optimization algorithm with the Branch-and-Bound Method and the Ant Colony Optimization to solve the multi-chromosomal reversal median problem. We convert the large-scale genome into TSP maps at first. Then we use a hybrid optimization algorithm with the Branch-and-Bound Method and the Ant Colony Optimization to solve the problem. In our improved algorithm, we increase the search speed by implement multi-branch parallel search of ACO. Our extensive experiments on simulated datasets show that this median solver is efficient.

关键词： Ant colony optimization

来源：评论

学校读者我要写书评

暂无评论

Improved cross-media relevance model for automatic image annotation

引用

ICIC Express Letters, Part B: Applications 2012年第5期3卷 1231-1237页

作者： Song, Haiyu Li, Xiongfei Wang, Pengjie Bao, Cuizhu Liu, Yongkui Xu, Lu Institute of Computer Graphics and Image Processing Dalian Nationalities University No. 18 Liaohe Road Economic and Technological Development Zone-Dalian 116600 China Key Laboratory of Symbol Computation and Knowledge Engineering Ministry of Education Jilin University No. 2699 Qianjin Avenue Changchun 130012 China

Automatic image annotation has been an active research topic in the last decade due to its potentially large impact on image retrieval, object recognition and image understanding. Many approaches have been proposed for image annotation in recent years. The most influential annotation model is relevance model, whose representative algorithm is Cross-Media Relevance Model (CMRM). In this paper, to overcome the shortcomings of the original CMRM, we propose improved CMRM. In the improved CMRM, the annotation word distribution is estimated using binary probability, the significance of blob is discriminately depicted using TF-IDF (term frequency-inverse document frequency) schemes, and the similarity measurement of blob is based on blob and continuousvalued feature vector. The experimental results on the standard Corel5K dataset show that our proposed approach outperforms the original CMRM by about 111% and 73% in terms of F1 and N+ measures respectively. The improved CMRM can get better performance in specific domains such as cultural heritage and special necessities digitization for ethnic minorities. © 2012 ICIC International.

关键词： Image retrieval

来源：评论

学校读者我要写书评

暂无评论

An approach for calculating semantic similarity between words using WordNet

An approach for calculating semantic similarity between word...

引用

2011 2nd International Conference on Digital Manufacturing and Automation, ICDMA 2011

作者： Hao, Dou Zuo, Wanli Peng, Tao He, Fengling College of Computer Science and Technology Jilin University Key Laboratory of Symbol Computation and Knowledge Engineering of the Ministry of Education Changchun China

ISBN: (纸本)9780769544557

This paper presents a new edge-counting based method using Word Net to compute the similarity. The method achieves a similarity that perfectly fits with human rating and effectively simulate the human tHought process that is people prefer to consider more differences when the semantic distance between two word is closer, and vice versa. At last, we weigh up our model against a benchmark set by human similarity judgment, and obtain a much improved result compared with other methods. © 2011 IEEE.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

A study on improved hidden Markov models and applications to speech recognition

A study on improved hidden Markov models and applications to...

引用

2011 International Conference on Computer Science and Service System, CSSS 2011

作者： Zhang, Zeliang Li, Xiongfei Key Laboratory of Symbol Computation and Knowledge Engineering Ministry of Education Jilin University Changchun China Beihua University Jilin China

ISBN: (纸本)9781424497638

Visual voice lip-reading, so the computer can understand what the speakers want to express direction by looking at their lips. Lip reading is the easiest way to compare the early characters and templates from the frozen image is stored. It ignores the very nature and time changes. This method is very simple, but it's just simple elements can be classified, then it may not show significant speech recognition services. Behavior was characterized by more and more common. Because of the hidden Markov model is superior (HMM), which can be widely used in speech recognition. In recent years, is also used to lip reading identification. Classical HMM model, so that the two assumptions: hidden assumptions collected: in t+1 the state can only be in this country is not in the state before t;from the hidden visible state hypothesis: only by regulating the t hide the visible state, rather than the previous state. This hypothesis is not very useful in some applications (such as lip reading) is reasonable. Under certain conditions, in the t state not only limits the t1, but also t2. Therefore, this study modified the assumptions of the classical HMM to derive a new HMM model and algorithms, and applied to the lip-reading recognition is increasing discrimination. © 2011 IEEE.

关键词： Hidden Markov models

来源：评论

学校读者我要写书评

暂无评论

Communities detection with applications to real-world networks analysis

Communities detection with applications to real-world networ...

引用

2011 7th International Conference on computational Intelligence and Security, CIS 2011

作者： Yang, Bo School of Computer Science and Technology Key Laboratory of Symbol Computation and Knowledge Engineering of the Ministry of Education Jilin University Changchun 130012 China

ISBN: (纸本)9780769545844

Community structure is one of non-trivial topological properties ubiquitously demonstrated in real-world complex networks. Related theories and approaches are of fundamental importance for understanding the functions of networks. Previously, we have proposed a probabilistic algorithm called the NCMA to efficiently as well as effectively mine communities from real-world networks. Here, we show that the NCMA can be readily extended and applied to address a wide range of network oriented applications beyond community detection including ranking, characterizing and searching real-world networks. © 2011 IEEE.

关键词： Complex networks

来源：评论

学校读者我要写书评

暂无评论

A new feature selection method based on clustering

A new feature selection method based on clustering

引用

2011 8th International Conference on Fuzzy Systems and knowledge Discovery, FSKD 2011, Jointly with the 2011 7th International Conference on Natural computation, ICNC'11

作者： Liu, Huawen Mo, Yuchang Wang, Jiyi Zhao, Jianmin Department of Computer Science Zhejiang Normal University Jinhua 321004 China Key Laboratory of Symbol Computation and Knowledge Engineering of Ministry of Education Changchun 130012 China School of Computer Science and Engineering Southeast University Nanjing 211189 China

ISBN: (纸本)9781612841816

Feature selection is an effective technique to put the high dimension of data down, which is prevailing in many application domains, such as text categorization and bio-informatics, and can bring many advantages, such as improving efficiency and avoiding over-fitting, to learning algorithms. Currently, many efforts have been attempted in this field and various feature selection methods have been developed and proved to be very competitive. Unlike other selection methods, in this paper we propose a new method to select important features using a manner of feature clustering. The main character of our method is that it works like data clustering in an agglomerative way. In this method, each feature is considered as a data point clustered with between-cluster and within-cluster distances. As a result, the selected feature subset has minimal redundancy among its members and maximal relevance with the class labels. Our performance evaluations on seven benchmark datasets show that the classification performance achieved by our proposed method is better than other feature selection methods. © 2011 IEEE.

关键词： Feature Selection

来源：评论

学校读者我要写书评

暂无评论

A novel method for RNA secondary structure prediction

A novel method for RNA secondary structure prediction

引用

2011 7th International Conference on Natural computation, ICNC 2011

作者： Xing, Chong Ji, Zhaohua Wang, Gang Wang, Yao Shen, Wei Liang, Yanchun Key Laboratory for Symbol Computation and Knowledge Engineering of National Education MinistryCollege of Computer Science and TechnologyJilin University Changchun China College of Information Engineering Guanghua College of Changchun University Changchun China College of Communications Engineering Jilin University Changchun China

ISBN: (纸本)9781424499533

In this paper, we propose PSOfold, a particle swarm optimization for RNA secondary structure prediction. PSOfold is based on the recently published IPSO. We present two strategies to improve the performance of IPSO. Firstly, in order to boost the competence in searching an optimal solution, fuzzy logic control is used to adaptively adjust the parameters in PSO. Accordingly, three fuzzy logic controls are designed by which the inertia weight, learning factors and the number of ants are tuned respectively. Secondly, to further settle the stem permutation problem, we put forward a solution conversion strategy (SCS), which can transform discrete values of stems into an ordered stem combination, thereby supplying an enhanced solution to evaluation of objective function. An evaluation of the performance of PSOfold in terms of prediction accuracy is made via comparison with one dynamic programming algorithm mfold and four metaheuristics, IPSO, ACRNA, RnaPredict, SARNA-Predict and mfold for ten individual known structures. PSOfold is able to predict structures with higher prediction accuracy than the other metaheuristic based methods on certain sequences, and has comparable performance compared with mfold. © 2011 IEEE.

关键词： Particle swarm optimization (PSO)

来源：评论

学校读者我要写书评

暂无评论

A Chinese online document clustering algorithm based on hidden sentiment vector

引用

Journal of computational Information Systems 2011年第6期7卷 1974-1981页

作者： Yang, Hui Wang, Zhe Zhou, Xu Wang, Zuo Zhou, Tong College of Computer Science and Technology Jilin University Changchun 130012 China Key Laboratory for Symbol Computation and Knowledge Engineering Ministry of Education China

The core idea of clustering algorithm is the division of data into groups of similar objects. Some clustering algorithms are proven good performance on document clustering, such as k-means and UPGMA etc. However, few document clustering algorithms pay attention to the hidden sentiment which is a very important feature of the documents. This paper presents an improved k-means algorithm (HSK-Means) based on hidden sentiment vector for Chinese document clustering. Chinese dependency grammar rules are used to extract document feature and evaluate hidden sentiment. A new method for selecting initial cluster centroids based on rank mechanism of similarity is also proposed to improve accuracy of clustering algorithm. The experimental results on real online document sets illustrate that the performance of HSK-Means algorithm is better than classic document clustering algorithms. Copyright © 2011 Binary Information Press.

关键词： Information retrieval

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：