Pub Date : 2009-12-04DOI: 10.1109/CCPR.2009.5344150
Lihua Guo, Lianwen Jin
Multi-instance and Multi-Label (MIML) machine learning has been employed in the generic object classification for it's gracefully performance in solving the ambiguity of image. The whole image is regarded as a multi-instance bag. The image is separated into four parts, whose edge's histograms are calculated. These input vectors can be combined a multi-instance ones for adapting the MIML learning. The experimental results show that the average precise ratio of our method is higher 3% than one of the traditional Support Vector Machine method.
{"title":"The Generic Object Classification Based on MIML Machine Learning","authors":"Lihua Guo, Lianwen Jin","doi":"10.1109/CCPR.2009.5344150","DOIUrl":"https://doi.org/10.1109/CCPR.2009.5344150","url":null,"abstract":"Multi-instance and Multi-Label (MIML) machine learning has been employed in the generic object classification for it's gracefully performance in solving the ambiguity of image. The whole image is regarded as a multi-instance bag. The image is separated into four parts, whose edge's histograms are calculated. These input vectors can be combined a multi-instance ones for adapting the MIML learning. The experimental results show that the average precise ratio of our method is higher 3% than one of the traditional Support Vector Machine method.","PeriodicalId":354468,"journal":{"name":"2009 Chinese Conference on Pattern Recognition","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124923506","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Identification of sentiment orientation in Chinese words is essential for getting sentiment comprehension of Chinese text, and building a basic semantic lexicon with Chinese emotional words will provide a core subset for identifying emotional words in a special area. It can not only help to identify and enlarge semantic lexicon in corpus effectively but also improve classification efficiency. On the basis of the similarity of Chinese words, the paper has proposed a method of calculating sentiment weight of Chinese emotional words. In addition, a dictionary with basic Chinese emotional words has been constructed based on the HowNet semantic lexicon. By utilizing the dictionary together with TF-IDF, we have done experiments to identify sentiment orientation in Chinese text and have got satisfying classification result.
{"title":"A Method of Building Chinese Basic Semantic Lexicon Based on Word Similarity","authors":"Yan-hui Zhu, Zhi-qiang Wen, Ping Wang, Zhao-yi Peng","doi":"10.1109/CCPR.2009.5344041","DOIUrl":"https://doi.org/10.1109/CCPR.2009.5344041","url":null,"abstract":"Identification of sentiment orientation in Chinese words is essential for getting sentiment comprehension of Chinese text, and building a basic semantic lexicon with Chinese emotional words will provide a core subset for identifying emotional words in a special area. It can not only help to identify and enlarge semantic lexicon in corpus effectively but also improve classification efficiency. On the basis of the similarity of Chinese words, the paper has proposed a method of calculating sentiment weight of Chinese emotional words. In addition, a dictionary with basic Chinese emotional words has been constructed based on the HowNet semantic lexicon. By utilizing the dictionary together with TF-IDF, we have done experiments to identify sentiment orientation in Chinese text and have got satisfying classification result.","PeriodicalId":354468,"journal":{"name":"2009 Chinese Conference on Pattern Recognition","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124943394","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-12-04DOI: 10.1109/CCPR.2009.5343954
Longbiao Wang, Yoshiki Kishi, A. Kai
Channel distortion for a distant environment may drastically degrade the performance of speaker recognition because the training and test conditions differ significantly. In this paper, we propose robust distant speaker recognition that is based on the automatic selection of reverberant environments using Gaussian mixture models. Three methods involving (I) optimum channel determination, (II) joint optimum speaker and channel determination, or (III) optimum channel determination at the utterance level are proposed. Real-world speech data and simulated reverberant speech data are used to evaluate our proposed methods. The third proposed method achieves a relative error reduction of 69.6% over (baseline) speaker recognition using a reverberant environment-independent method, and it has performance equivalent to that of a
{"title":"Distant Speaker Recognition Based on the Automatic Selection of Reverberant Environments Using GMMs","authors":"Longbiao Wang, Yoshiki Kishi, A. Kai","doi":"10.1109/CCPR.2009.5343954","DOIUrl":"https://doi.org/10.1109/CCPR.2009.5343954","url":null,"abstract":"Channel distortion for a distant environment may drastically degrade the performance of speaker recognition because the training and test conditions differ significantly. In this paper, we propose robust distant speaker recognition that is based on the automatic selection of reverberant environments using Gaussian mixture models. Three methods involving (I) optimum channel determination, (II) joint optimum speaker and channel determination, or (III) optimum channel determination at the utterance level are proposed. Real-world speech data and simulated reverberant speech data are used to evaluate our proposed methods. The third proposed method achieves a relative error reduction of 69.6% over (baseline) speaker recognition using a reverberant environment-independent method, and it has performance equivalent to that of a","PeriodicalId":354468,"journal":{"name":"2009 Chinese Conference on Pattern Recognition","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123837875","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-12-04DOI: 10.1109/CCPR.2009.5344079
Ziqiang Wang, Xia Sun
To efficiently deal with the curse of dimensionality in the content-based image retrieval(CBIR) system, a novel image retrieval algorithm is proposed by combination of local discriminant embedding(LDE) and least square SVM(LS-SVM) in this paper. LDE aims to achieve good discriminating performance by integrating the local geometrical structure and class relations between image data. LS-SVM classifier is used to classify the retrieved image into relevant or irrelevant image based on extracted low-level visual features. Experimental results on real-world image collection demonstrate that the proposed algorithm performs much better than other related image retrieval algorithms.
{"title":"Image Retrieval Using Discriminant Embedding and LS-SVM","authors":"Ziqiang Wang, Xia Sun","doi":"10.1109/CCPR.2009.5344079","DOIUrl":"https://doi.org/10.1109/CCPR.2009.5344079","url":null,"abstract":"To efficiently deal with the curse of dimensionality in the content-based image retrieval(CBIR) system, a novel image retrieval algorithm is proposed by combination of local discriminant embedding(LDE) and least square SVM(LS-SVM) in this paper. LDE aims to achieve good discriminating performance by integrating the local geometrical structure and class relations between image data. LS-SVM classifier is used to classify the retrieved image into relevant or irrelevant image based on extracted low-level visual features. Experimental results on real-world image collection demonstrate that the proposed algorithm performs much better than other related image retrieval algorithms.","PeriodicalId":354468,"journal":{"name":"2009 Chinese Conference on Pattern Recognition","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123942699","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-12-04DOI: 10.1109/CCPR.2009.5344151
Yinfeng Luo, Shenglin Yu
Particle filter is an effective method for non-linear filter and it has been gained special attention of researchers in various fields. There will be a new mixed particle filter (PUPF) proposed in this paper based on the general particle filter and the unscented particle filter. lt first uses the general particle filter to generate particles for estimating the state at time k and then a new auxiliary model will be introduced. We would use the unscented particle filter to estimate the state at time k the second time. This structure makes use of the latest observation information, it has small error and better stability. The experimental results indicate that the proposed particle filter's performance outperforms the other four particle filters .The result indicates that the PUPF is a useful method for nonlinear filter problems.
{"title":"A New Mixed Particle Filter Based on an Auxiliary Model","authors":"Yinfeng Luo, Shenglin Yu","doi":"10.1109/CCPR.2009.5344151","DOIUrl":"https://doi.org/10.1109/CCPR.2009.5344151","url":null,"abstract":"Particle filter is an effective method for non-linear filter and it has been gained special attention of researchers in various fields. There will be a new mixed particle filter (PUPF) proposed in this paper based on the general particle filter and the unscented particle filter. lt first uses the general particle filter to generate particles for estimating the state at time k and then a new auxiliary model will be introduced. We would use the unscented particle filter to estimate the state at time k the second time. This structure makes use of the latest observation information, it has small error and better stability. The experimental results indicate that the proposed particle filter's performance outperforms the other four particle filters .The result indicates that the PUPF is a useful method for nonlinear filter problems.","PeriodicalId":354468,"journal":{"name":"2009 Chinese Conference on Pattern Recognition","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125171618","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-12-04DOI: 10.1109/CCPR.2009.5343984
Sumuya Bao, Chonghui Guo, Shanglei Chai
Recently spectral clustering has become one of the most popular clustering algorithms. Although it has many advantages, it still has a lot of shortcomings which should be resolved, such as there are a wide variety of spectral clustering algorithms that use the eigenvectors in slightly different ways and many of these algorithms have no proof that they will actually compute a reasonable clustering. The spectral clustering method based on normalized cut criterion is a very efficient spectral clustering method. In this paper, we give a note on why we choose the first k eigenvectors in the algorithm (rationality of the clustering) and the conditions for indicator vectors under which the clustering problem could lead to the problem of minimizing the objective function of the spectral clustering method based on normalized cut criterion.
{"title":"A Note on Spectral Clustering Method Based on Normalized Cut Criterion","authors":"Sumuya Bao, Chonghui Guo, Shanglei Chai","doi":"10.1109/CCPR.2009.5343984","DOIUrl":"https://doi.org/10.1109/CCPR.2009.5343984","url":null,"abstract":"Recently spectral clustering has become one of the most popular clustering algorithms. Although it has many advantages, it still has a lot of shortcomings which should be resolved, such as there are a wide variety of spectral clustering algorithms that use the eigenvectors in slightly different ways and many of these algorithms have no proof that they will actually compute a reasonable clustering. The spectral clustering method based on normalized cut criterion is a very efficient spectral clustering method. In this paper, we give a note on why we choose the first k eigenvectors in the algorithm (rationality of the clustering) and the conditions for indicator vectors under which the clustering problem could lead to the problem of minimizing the objective function of the spectral clustering method based on normalized cut criterion.","PeriodicalId":354468,"journal":{"name":"2009 Chinese Conference on Pattern Recognition","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128226549","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-12-04DOI: 10.1109/CCPR.2009.5343977
Toshinori Miyoshi, T. Nagasaki, Hiroshi Shinjo
Normalization is a particular important preprocessing operation, and has a large effect on the performance of character recognition. One of the purposes of normalization is to regulate the size, position, and shape of character images so as to reduce within-class shape variations. Among various methods of normalization, moment-based normalizations are known to greatly improve the performance of character recognition. However, conventional moment-based normalization methods are susceptible to the variations of stroke length and/or thickness. In order to alleviate this problem, we propose moment normalization methods that use the moments of character contours instead of character images themselves to estimate the transformation parameters. Our experiments show that the proposed methods are effective particularly for printed character recognition.
{"title":"Character Normalization Methods Using Moments of Gradient Features and Normalization Cooperated Feature Extraction","authors":"Toshinori Miyoshi, T. Nagasaki, Hiroshi Shinjo","doi":"10.1109/CCPR.2009.5343977","DOIUrl":"https://doi.org/10.1109/CCPR.2009.5343977","url":null,"abstract":"Normalization is a particular important preprocessing operation, and has a large effect on the performance of character recognition. One of the purposes of normalization is to regulate the size, position, and shape of character images so as to reduce within-class shape variations. Among various methods of normalization, moment-based normalizations are known to greatly improve the performance of character recognition. However, conventional moment-based normalization methods are susceptible to the variations of stroke length and/or thickness. In order to alleviate this problem, we propose moment normalization methods that use the moments of character contours instead of character images themselves to estimate the transformation parameters. Our experiments show that the proposed methods are effective particularly for printed character recognition.","PeriodicalId":354468,"journal":{"name":"2009 Chinese Conference on Pattern Recognition","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130393936","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-12-04DOI: 10.1109/CCPR.2009.5344071
Wenjia Yang, L. Dou, Juan Zhang
To solve the problem of moving object segmentation in video sequence, a new video moving object segmentation algorithm was proposed based on Kirsch edge operator. The detected edge is mainly analyzed for segmentation and motion vector field is taken as assistant information. Firstly, the motion vectors are processed by accumulation and median filer. Secondly, templates of Kirsch operators are decomposed into difference templates and common templates to find the edge position; then, the edge information and the motion vectors are fused to get moving object by adaptive state labeling. The experimental results show the proposed algorithm has a better veracity of segmentation.
{"title":"Video Moving Object Segmentation Algorithm Based on an Improved Kirsch Edge Operator","authors":"Wenjia Yang, L. Dou, Juan Zhang","doi":"10.1109/CCPR.2009.5344071","DOIUrl":"https://doi.org/10.1109/CCPR.2009.5344071","url":null,"abstract":"To solve the problem of moving object segmentation in video sequence, a new video moving object segmentation algorithm was proposed based on Kirsch edge operator. The detected edge is mainly analyzed for segmentation and motion vector field is taken as assistant information. Firstly, the motion vectors are processed by accumulation and median filer. Secondly, templates of Kirsch operators are decomposed into difference templates and common templates to find the edge position; then, the edge information and the motion vectors are fused to get moving object by adaptive state labeling. The experimental results show the proposed algorithm has a better veracity of segmentation.","PeriodicalId":354468,"journal":{"name":"2009 Chinese Conference on Pattern Recognition","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121313098","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-12-04DOI: 10.1109/CCPR.2009.5344074
Yu Song, Qingling Li, F. Sun
Shannon entropy is effective uncertainty measurement criterion for stochastic system. In this paper, adaptive fusion particle filter is proposed for visual tracking by introduced Shannon entropy in particle filter framework. Firstly, the particle filter, which is considered as the process of particles assimilating negative entropy to reduce uncertainty, is surveyed from viewpoint of information theory. Secondly, maximum negative entropy criterion is proposed to select tracking feature form features pool online. At last, color histogram and edge orientation histogram features are utilized in experiments, tracking results show that the proposed algorithm is a robust and accuracy tracking algorithm.
{"title":"Shannon Entropy-Based Adaptive Fusion Particle Filter for Visual Tracking","authors":"Yu Song, Qingling Li, F. Sun","doi":"10.1109/CCPR.2009.5344074","DOIUrl":"https://doi.org/10.1109/CCPR.2009.5344074","url":null,"abstract":"Shannon entropy is effective uncertainty measurement criterion for stochastic system. In this paper, adaptive fusion particle filter is proposed for visual tracking by introduced Shannon entropy in particle filter framework. Firstly, the particle filter, which is considered as the process of particles assimilating negative entropy to reduce uncertainty, is surveyed from viewpoint of information theory. Secondly, maximum negative entropy criterion is proposed to select tracking feature form features pool online. At last, color histogram and edge orientation histogram features are utilized in experiments, tracking results show that the proposed algorithm is a robust and accuracy tracking algorithm.","PeriodicalId":354468,"journal":{"name":"2009 Chinese Conference on Pattern Recognition","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-12-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121320504","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：