The problem of highlight detection and recognition in soccer video has been a hot research topic in many fields, such as image processing, pattern recognition, and machine learning. This paper puts forward a novel algorithm to detect soccer highlights, which is made up of three phases. In the first stage, slow motion replay segments are located through the analysis of visual rhythm and the structure tensor histogram; In the second stage, we concentrate on the detection of the goal net according to its edge projection in the vertical direction; In the end, a set of heuristic rules are employed to identify the goal shots or near-miss shots from the soccer video based on the recognition result of the previous stage. Experimental results show that our method is effective and efficient.
{"title":"Study on Highlights Detection in Soccer Video Based on the Location of Slow Motion Replay and Goal Net Recognition","authors":"X. Ruan, Shijin Li, Yan Dong, Jun Feng","doi":"10.1109/CCPR.2008.41","DOIUrl":"https://doi.org/10.1109/CCPR.2008.41","url":null,"abstract":"The problem of highlight detection and recognition in soccer video has been a hot research topic in many fields, such as image processing, pattern recognition, and machine learning. This paper puts forward a novel algorithm to detect soccer highlights, which is made up of three phases. In the first stage, slow motion replay segments are located through the analysis of visual rhythm and the structure tensor histogram; In the second stage, we concentrate on the detection of the goal net according to its edge projection in the vertical direction; In the end, a set of heuristic rules are employed to identify the goal shots or near-miss shots from the soccer video based on the recognition result of the previous stage. Experimental results show that our method is effective and efficient.","PeriodicalId":292956,"journal":{"name":"2008 Chinese Conference on Pattern Recognition","volume":"153 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115188266","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}
Face recognition, which is an active research area in pattern recognition, has made great progress in recent years. Its performance that based on multiple face images is satisfying, but it is remain poor when only a single face image is used to training. Accordingly, we propose a new algorithm of face recognition that based on a single face image in this paper. The new algorithm can be divided into three steps: first, we compute horizontal and vertical edge images from the gray image; then, local binary pattern histogram is extracted from those two edge images; finally, elastic matching is used to classification. Experimental result on some standard face databases show that our proposed method can substantially improves the recognition performance and is robustness to pose, illumination and expression.
{"title":"A Novel Algorithm for Face Recognition Based on a Single Image","authors":"NuTao Tan, Lei Huang, Chang-ping Liu","doi":"10.1109/CCPR.2008.54","DOIUrl":"https://doi.org/10.1109/CCPR.2008.54","url":null,"abstract":"Face recognition, which is an active research area in pattern recognition, has made great progress in recent years. Its performance that based on multiple face images is satisfying, but it is remain poor when only a single face image is used to training. Accordingly, we propose a new algorithm of face recognition that based on a single face image in this paper. The new algorithm can be divided into three steps: first, we compute horizontal and vertical edge images from the gray image; then, local binary pattern histogram is extracted from those two edge images; finally, elastic matching is used to classification. Experimental result on some standard face databases show that our proposed method can substantially improves the recognition performance and is robustness to pose, illumination and expression.","PeriodicalId":292956,"journal":{"name":"2008 Chinese Conference on Pattern Recognition","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123463813","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}
Evaluation of Chinese handwriting character quality is an important function of computer assisted Chinese learning technology it can point out the errors in a handwritten character and make objective assessment on the writing quality of the character. However, only a few studies were reported on this new research topic in the literature. In this paper, the main target of handwriting evaluation has been presented. The common layout errors in handwriting samples are summarized and then a new layout evaluation method is proposed. The method is consisted of three parts of stroke layout evaluation, component layout evaluation and entire character shape evaluation, through the application of nine assessment rules. The experiment results show that the method is capable for detecting layout errors of handwriting samples and making objective assessment on whether a character is written good or not.
{"title":"A Method for Layout Evaluation of Online Handwritten Chinese Character Quality Based on Template","authors":"Weiping Xia, Lianwen Jin","doi":"10.1109/CCPR.2008.75","DOIUrl":"https://doi.org/10.1109/CCPR.2008.75","url":null,"abstract":"Evaluation of Chinese handwriting character quality is an important function of computer assisted Chinese learning technology it can point out the errors in a handwritten character and make objective assessment on the writing quality of the character. However, only a few studies were reported on this new research topic in the literature. In this paper, the main target of handwriting evaluation has been presented. The common layout errors in handwriting samples are summarized and then a new layout evaluation method is proposed. The method is consisted of three parts of stroke layout evaluation, component layout evaluation and entire character shape evaluation, through the application of nine assessment rules. The experiment results show that the method is capable for detecting layout errors of handwriting samples and making objective assessment on whether a character is written good or not.","PeriodicalId":292956,"journal":{"name":"2008 Chinese Conference on Pattern Recognition","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126407100","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}
It is the key to improve the natural degree of speech synthesis and reduce the error rate of speech recognition that analyzes the information structure and prosodic structure of sentence and chapters. Based on large speech corpus (ASCCD) with prosodic structure label, we measured the characteristics of duration and pitch on prosodic phrase. The statistical results on duration and pitch are presented in this paper.(1)The prosodic border can obviously prolong the duration of syllable, different tone and accent have different effect to prolong the syllable duration.(2)The break duration at prosodic border, especially at little prosodic border is more obvious. It is obvious that F0 reset always occurs between prosodic phrases. The F0 bottom line is always declined. The F0 top line is declined after the accent. And at accent position, the rage of pitch is big and the top line is high.
{"title":"Durational Characteristics and Pitch Characteristics of the Prosodic Phrase in Mandarin Chinese","authors":"Chongjia Ni, Wenju Liu","doi":"10.1109/CCPR.2008.84","DOIUrl":"https://doi.org/10.1109/CCPR.2008.84","url":null,"abstract":"It is the key to improve the natural degree of speech synthesis and reduce the error rate of speech recognition that analyzes the information structure and prosodic structure of sentence and chapters. Based on large speech corpus (ASCCD) with prosodic structure label, we measured the characteristics of duration and pitch on prosodic phrase. The statistical results on duration and pitch are presented in this paper.(1)The prosodic border can obviously prolong the duration of syllable, different tone and accent have different effect to prolong the syllable duration.(2)The break duration at prosodic border, especially at little prosodic border is more obvious. It is obvious that F0 reset always occurs between prosodic phrases. The F0 bottom line is always declined. The F0 top line is declined after the accent. And at accent position, the rage of pitch is big and the top line is high.","PeriodicalId":292956,"journal":{"name":"2008 Chinese Conference on Pattern Recognition","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128940418","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}
This paper presents an approach for compression and synthesis of BTF (bidirectional texture function), and using it BTF can be mapped on arbitrary surfaces. BTF is a kind of multidimensional texture data, which can perfectly shows the self-shadow, self-occlusion and inter-reflection reflectance of object surfaces under varying light directions. However, it is difficult to use BTF on object surfaces because of its enormous data and small size. Our method uses PCA algorithm to compress enormous BTF data, and we propose a synthesis algorithm to solve its size problem and insure multidimensional consistency, while we also propose an algorithm of error calculation for Wang tiles' construction. Using the method we proposed this paper we can rendering BTF on arbitrary object surfaces efficiently and quickly.
{"title":"An Approach for Compression and Synthesis of BTF","authors":"Zhan Zhang, Yue Qi, Yong Hu","doi":"10.1109/CCPR.2008.38","DOIUrl":"https://doi.org/10.1109/CCPR.2008.38","url":null,"abstract":"This paper presents an approach for compression and synthesis of BTF (bidirectional texture function), and using it BTF can be mapped on arbitrary surfaces. BTF is a kind of multidimensional texture data, which can perfectly shows the self-shadow, self-occlusion and inter-reflection reflectance of object surfaces under varying light directions. However, it is difficult to use BTF on object surfaces because of its enormous data and small size. Our method uses PCA algorithm to compress enormous BTF data, and we propose a synthesis algorithm to solve its size problem and insure multidimensional consistency, while we also propose an algorithm of error calculation for Wang tiles' construction. Using the method we proposed this paper we can rendering BTF on arbitrary object surfaces efficiently and quickly.","PeriodicalId":292956,"journal":{"name":"2008 Chinese Conference on Pattern Recognition","volume":"76 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127484311","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}
Small sample size of training data might bring trouble as the bias of the estimated parameters for a pattern recognition system. Plug-in test statistics suffer from large estimation errors, often causing the performance to degrade as the measurement vector dimension increases. The informative component extraction method helps to solve this problem by throwing out some dimensions which have relative small distance to the nominal model in statistic sense. Preserving the discriminative information for identification increases the performance. Considering the distortion of the estimated distribution, we introduce the idea of robustness in the informative component extraction. A tolerance ball is applied in the selection of informative and robust components for each individual model (hypothesis). When dealing with multiple parameters model, the supreme of all tolerance balls is used. Informative component extraction with robustness consideration could be used in nonparametric density case simply with slight modification. We use two methods to extract informative component and the performance is examined with 4 different training data sets. The simulation results are compared and discussed with improved performance when considering the robustness.
{"title":"Informative Component Extraction with Robustness Consideration","authors":"Mei Chen, Yan Liu","doi":"10.1109/CCPR.2008.18","DOIUrl":"https://doi.org/10.1109/CCPR.2008.18","url":null,"abstract":"Small sample size of training data might bring trouble as the bias of the estimated parameters for a pattern recognition system. Plug-in test statistics suffer from large estimation errors, often causing the performance to degrade as the measurement vector dimension increases. The informative component extraction method helps to solve this problem by throwing out some dimensions which have relative small distance to the nominal model in statistic sense. Preserving the discriminative information for identification increases the performance. Considering the distortion of the estimated distribution, we introduce the idea of robustness in the informative component extraction. A tolerance ball is applied in the selection of informative and robust components for each individual model (hypothesis). When dealing with multiple parameters model, the supreme of all tolerance balls is used. Informative component extraction with robustness consideration could be used in nonparametric density case simply with slight modification. We use two methods to extract informative component and the performance is examined with 4 different training data sets. The simulation results are compared and discussed with improved performance when considering the robustness.","PeriodicalId":292956,"journal":{"name":"2008 Chinese Conference on Pattern Recognition","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132409578","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}
Due to the quality of service issues that wireless communications for data transmission need to ensure, we use the SOM neural network for QoS pattern space convergence, and apply cluster results to the shortest path algorithm in sensor networks. This paper constructs a wireless sensor networks packet loss rate test model of Simulink, and uses it to measure the packet loss rate in different communication distance and the noise power density. From these we obtain the SOM network input samples. After network training, the convergent vector matrix and the corresponding quality of service function are obtained. Finally, we apply the quality of service to the shortest path tree structure, and evaluate the performance of pattern recognition in the shortest path tree structure by NS2 software.
{"title":"Study on SOM in Wireless Sensor Networks QoS Measurement","authors":"G. Wang, S. Zhang","doi":"10.1109/CCPR.2008.95","DOIUrl":"https://doi.org/10.1109/CCPR.2008.95","url":null,"abstract":"Due to the quality of service issues that wireless communications for data transmission need to ensure, we use the SOM neural network for QoS pattern space convergence, and apply cluster results to the shortest path algorithm in sensor networks. This paper constructs a wireless sensor networks packet loss rate test model of Simulink, and uses it to measure the packet loss rate in different communication distance and the noise power density. From these we obtain the SOM network input samples. After network training, the convergent vector matrix and the corresponding quality of service function are obtained. Finally, we apply the quality of service to the shortest path tree structure, and evaluate the performance of pattern recognition in the shortest path tree structure by NS2 software.","PeriodicalId":292956,"journal":{"name":"2008 Chinese Conference on Pattern Recognition","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128098857","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}
After analysis and comparison of the problems of the existing one-versus-one (OVO) and one-versus-rest (OVR) decomposition methods of multi-class support vector machine (SVM), the novel strategy based on posterior probability is presented to reconstruct a multi-class classifier from binary SVM-based classifiers. The new reconstruction strategy can increase recognition accuracy and resolve the unclassifiable region problems in the conventional ones. Firstly, the geometric distance of test sample to the optimal classification hyperplane is used as the criterion of estimating the class probabilities to decrease the incomparability existing in different binary SVM-based classifiers. Then based on the Bayesian posterior probability theory, the combination strategy of the probability output among these binary SVM-based classifiers in OVO decomposition is given and the different prior probabilities of them are considered. Lastly, the prior probabilities are evaluated by OVR decomposition. In order to verify the effectiveness of this strategy, experiments have been made on UCI database; the experiment results show that the reconstruction strategy presented is effective over conventional ones.
{"title":"Reconstruction Strategy for Multi-Class SVM Based on Posterior Probability","authors":"Deihui Wu","doi":"10.1109/CCPR.2008.21","DOIUrl":"https://doi.org/10.1109/CCPR.2008.21","url":null,"abstract":"After analysis and comparison of the problems of the existing one-versus-one (OVO) and one-versus-rest (OVR) decomposition methods of multi-class support vector machine (SVM), the novel strategy based on posterior probability is presented to reconstruct a multi-class classifier from binary SVM-based classifiers. The new reconstruction strategy can increase recognition accuracy and resolve the unclassifiable region problems in the conventional ones. Firstly, the geometric distance of test sample to the optimal classification hyperplane is used as the criterion of estimating the class probabilities to decrease the incomparability existing in different binary SVM-based classifiers. Then based on the Bayesian posterior probability theory, the combination strategy of the probability output among these binary SVM-based classifiers in OVO decomposition is given and the different prior probabilities of them are considered. Lastly, the prior probabilities are evaluated by OVR decomposition. In order to verify the effectiveness of this strategy, experiments have been made on UCI database; the experiment results show that the reconstruction strategy presented is effective over conventional ones.","PeriodicalId":292956,"journal":{"name":"2008 Chinese Conference on Pattern Recognition","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134053136","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}
Context-dependent acoustic model based on decision tree has been deeply investigated and applied in English, Chinese speech recognition. But in Mongolian speech recognition, little attention was paid in the past. In this paper Mongolian context-dependent acoustic model based on decision tree was proposed.and decision tree based state tying was applied to the acoustic model designning in Mongolian speech recognition. Finally, the experimental analysis was carries on to unseen triphone, sparse triphone by HTK platform and satisfactory effect is gained.
{"title":"Researching of Speech Recognition Oriented Mongolian Acoustic Model","authors":"Hasi Qilao, Guang-Lai Gao","doi":"10.1109/CCPR.2008.85","DOIUrl":"https://doi.org/10.1109/CCPR.2008.85","url":null,"abstract":"Context-dependent acoustic model based on decision tree has been deeply investigated and applied in English, Chinese speech recognition. But in Mongolian speech recognition, little attention was paid in the past. In this paper Mongolian context-dependent acoustic model based on decision tree was proposed.and decision tree based state tying was applied to the acoustic model designning in Mongolian speech recognition. Finally, the experimental analysis was carries on to unseen triphone, sparse triphone by HTK platform and satisfactory effect is gained.","PeriodicalId":292956,"journal":{"name":"2008 Chinese Conference on Pattern Recognition","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131936087","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}
A main problem in face retrieval is the semantic gap between low-level features and high-level semantic concepts. Relevance feedback (RF) may be used to incorporate to reduce the semantic gap. However, in the search for a specific target in a facial image database, a user's assignment of RF instances may be mistaken. This would make the system prediction of the user's target in a wrong way. Addressing this problem, we propose a new query point movement technique for target search by posing the problem of reducing the impact of inaccurate user feedback as an optimization problem. We develop a support vector machine based method to learn a decision boundary to identify ideal irrelevant images. Then we propose a rank function for finding target images, which would assign high scores to the images near the relevant images and punish those close to the decision boundary. Experiments are performed to show the stability and efficiency of the proposed algorithm.
{"title":"Reducing Impact of Inaccurate User Feedback in Face Retrieval","authors":"R. He, Weishi Zheng, Meng Ao, Stan Z. Li","doi":"10.1109/CCPR.2008.50","DOIUrl":"https://doi.org/10.1109/CCPR.2008.50","url":null,"abstract":"A main problem in face retrieval is the semantic gap between low-level features and high-level semantic concepts. Relevance feedback (RF) may be used to incorporate to reduce the semantic gap. However, in the search for a specific target in a facial image database, a user's assignment of RF instances may be mistaken. This would make the system prediction of the user's target in a wrong way. Addressing this problem, we propose a new query point movement technique for target search by posing the problem of reducing the impact of inaccurate user feedback as an optimization problem. We develop a support vector machine based method to learn a decision boundary to identify ideal irrelevant images. Then we propose a rank function for finding target images, which would assign high scores to the images near the relevant images and punish those close to the decision boundary. Experiments are performed to show the stability and efficiency of the proposed algorithm.","PeriodicalId":292956,"journal":{"name":"2008 Chinese Conference on Pattern Recognition","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133695910","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：