Nowadays digital watermarking has played an important role in the copyright protection of multimedia. A new robust watermarking algorithm is proposed with the base of error correcting code (ECC) and data fusion by evidence theory and distortion factor. The algorithm does not need any mask or original image during the extraction and makes the watermark more robust and further more the data fusion composed by a new watermark confidence relatively reduces the volume of the embedding watermark. At last experiments have proved the algorithm is practical.
{"title":"A Novel Watermarking Extraction Based on Error Correcting Code and Evidence Theory","authors":"Liping Chen, Zhiqiang Yao","doi":"10.1109/ICNC.2008.482","DOIUrl":"https://doi.org/10.1109/ICNC.2008.482","url":null,"abstract":"Nowadays digital watermarking has played an important role in the copyright protection of multimedia. A new robust watermarking algorithm is proposed with the base of error correcting code (ECC) and data fusion by evidence theory and distortion factor. The algorithm does not need any mask or original image during the extraction and makes the watermark more robust and further more the data fusion composed by a new watermark confidence relatively reduces the volume of the embedding watermark. At last experiments have proved the algorithm is practical.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"20 1","pages":"613-617"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88556504","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}
Discrete Hopfield neural network with delay is an extension of discrete Hopfield neural network. The dynamic behavior of discrete Hopfield neural networks with delay is mainly studied by using the method of defining energy function. The conditions for the networks with delay converging towards a limit cycle with length at most 2 and the conditions for the networks with delay converging towards a limit cycle with length 4 are respectively given. The obtained results here extend the existing results on stability of discrete Hopfield neural network with delay and without delay.
{"title":"Stability for Discrete Hopfield Neural Networks with Delay","authors":"H. Gao, J. Zhang, W. Qin","doi":"10.1109/ICNC.2008.303","DOIUrl":"https://doi.org/10.1109/ICNC.2008.303","url":null,"abstract":"Discrete Hopfield neural network with delay is an extension of discrete Hopfield neural network. The dynamic behavior of discrete Hopfield neural networks with delay is mainly studied by using the method of defining energy function. The conditions for the networks with delay converging towards a limit cycle with length at most 2 and the conditions for the networks with delay converging towards a limit cycle with length 4 are respectively given. The obtained results here extend the existing results on stability of discrete Hopfield neural network with delay and without delay.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"6 1","pages":"560-563"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87269536","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}
Although many significant achievements have been made on using ant colony optimization (ACO) algorithm to solve traveling salesman problem (TSP) and similar large-scale computational problems, the long convergent time required in the large-scale optimization still remains a computing bottle neck of ACO algorithm. In this paper, we present a rapidly convergent ant colony optimization (rcACO) algorithm to solve the TSP. In this algorithm, adaptive pheromone update is carried out according to the distance ants have moved, meanwhile, the inversion operator is used to enhance local search, etc. Our huge numerical experimental results demonstrate that the convergence speed of rcACO is tens to hundreds times faster than the recently improved ACO algorithms, meanwhile the global optimal solution can be achieved.
{"title":"An Efficient Approach for Solving TSP: The Rapidly Convergent Ant Colony Algorithm","authors":"Lingling Wang, Qingbao Zhu","doi":"10.1109/ICNC.2008.186","DOIUrl":"https://doi.org/10.1109/ICNC.2008.186","url":null,"abstract":"Although many significant achievements have been made on using ant colony optimization (ACO) algorithm to solve traveling salesman problem (TSP) and similar large-scale computational problems, the long convergent time required in the large-scale optimization still remains a computing bottle neck of ACO algorithm. In this paper, we present a rapidly convergent ant colony optimization (rcACO) algorithm to solve the TSP. In this algorithm, adaptive pheromone update is carried out according to the distance ants have moved, meanwhile, the inversion operator is used to enhance local search, etc. Our huge numerical experimental results demonstrate that the convergence speed of rcACO is tens to hundreds times faster than the recently improved ACO algorithms, meanwhile the global optimal solution can be achieved.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"22 1","pages":"448-452"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87292913","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}
We propose a novel, convenient way for the building of emotion thesaurus which can be used in assessing the affective qualities of natural languages contained in text. Our main goals are fast analysis and visualization of affective content for machines to communicate smoothly with humans and to realize emotion communications. Although there have been some studies about analyzing affective content in text, our primary unique method is mainly according to the main sememe of HowNet which is an on-line common sense knowledge base unveiling inter-conceptual relations and inter-attribute relations of concepts as connoting in lexicons of the Chinese and their English equivalents. Therefore our processing of the affective content is lead into the semantic level.
{"title":"The Building of Chinese Emotion Thesaurus Using HowNet Based on the Main Sememe","authors":"Bin Liu, F. Ren, Cong Wang","doi":"10.1109/ICNC.2008.194","DOIUrl":"https://doi.org/10.1109/ICNC.2008.194","url":null,"abstract":"We propose a novel, convenient way for the building of emotion thesaurus which can be used in assessing the affective qualities of natural languages contained in text. Our main goals are fast analysis and visualization of affective content for machines to communicate smoothly with humans and to realize emotion communications. Although there have been some studies about analyzing affective content in text, our primary unique method is mainly according to the main sememe of HowNet which is an on-line common sense knowledge base unveiling inter-conceptual relations and inter-attribute relations of concepts as connoting in lexicons of the Chinese and their English equivalents. Therefore our processing of the affective content is lead into the semantic level.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"47 1","pages":"91-95"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80603644","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}
Support vector machine (SVM) is an effective algorithm in pattern recognition. But usually, standard SVM requires solving a quadratic program (QP) problem. In majority situations, most implementations of SVM are approximate solution to the QP problem. As the approximate solutions cannot achieve the expected performance of SRM theory, it is necessary to research ensemble methods for SVM. Recently, in order to augment the diversities of individual classifiers of SVM, many researchers use random partition with the whole training to form sub-training sets. Therefore the performance of aggregated SVM, which was trained on those subsets, was improved. We proposed the ensemble method based on different implementations of SVM, because they have large diversities by their different implementing methods. The experiment results showed that this method is effectively to improve the aggregated learner's performance.
{"title":"Ensemble Implementations on Diversified Support Vector Machines","authors":"Kunlun Li, Yun-Long Dai, Wei Zhang","doi":"10.1109/ICNC.2008.197","DOIUrl":"https://doi.org/10.1109/ICNC.2008.197","url":null,"abstract":"Support vector machine (SVM) is an effective algorithm in pattern recognition. But usually, standard SVM requires solving a quadratic program (QP) problem. In majority situations, most implementations of SVM are approximate solution to the QP problem. As the approximate solutions cannot achieve the expected performance of SRM theory, it is necessary to research ensemble methods for SVM. Recently, in order to augment the diversities of individual classifiers of SVM, many researchers use random partition with the whole training to form sub-training sets. Therefore the performance of aggregated SVM, which was trained on those subsets, was improved. We proposed the ensemble method based on different implementations of SVM, because they have large diversities by their different implementing methods. The experiment results showed that this method is effectively to improve the aggregated learner's performance.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"26 1","pages":"180-184"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91224396","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}
Coverage is one of the main problems to be solved for wireless sensor networks (WSN). In some monitoring regions, the condition is very bad and worse cases often suddenly occur, the nodes of wireless sensor network need to dynamically change their position quickly and automatically re-coverage according to the monitoring events to achieve better monitoring results. The current algorithms are often limited to realize the optimal coverage of fixed region. Combined with artificial neural network, putting the improved growing neural gas with utility criterion algorithm into wireless sensor network, the network can rapid re-coverage with respond to the changed region especially for special environments. In order to speed the learning procedure, we use GA and SA which combines the ability of evolution of GA and probability searching of SA. The simulation results show that, compared with growing neural gas algorithm, growing neural gas with utility criterion algorithm and improved GNG algorithm, the improve GNG-U algorithm can reduce a lot of redundant nodes, improve mobility of the network, accelerate the rate of convergence and arrive optimal re-coverage.
{"title":"Hybrid Learning Algorithm for Effective Coverage in Wireless Sensor Networks","authors":"Yanjing Sun, Li Li","doi":"10.1109/ICNC.2008.320","DOIUrl":"https://doi.org/10.1109/ICNC.2008.320","url":null,"abstract":"Coverage is one of the main problems to be solved for wireless sensor networks (WSN). In some monitoring regions, the condition is very bad and worse cases often suddenly occur, the nodes of wireless sensor network need to dynamically change their position quickly and automatically re-coverage according to the monitoring events to achieve better monitoring results. The current algorithms are often limited to realize the optimal coverage of fixed region. Combined with artificial neural network, putting the improved growing neural gas with utility criterion algorithm into wireless sensor network, the network can rapid re-coverage with respond to the changed region especially for special environments. In order to speed the learning procedure, we use GA and SA which combines the ability of evolution of GA and probability searching of SA. The simulation results show that, compared with growing neural gas algorithm, growing neural gas with utility criterion algorithm and improved GNG algorithm, the improve GNG-U algorithm can reduce a lot of redundant nodes, improve mobility of the network, accelerate the rate of convergence and arrive optimal re-coverage.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"12 1","pages":"227-231"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91298413","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}
In this paper, the solvability conditions and the general solution of matrix equations (AX=B,XC=D) with a submatrix constraint are obtained by using the SVD(singular value decomposition) of matrix. In addition, the expression of the optimal approximation solution to a given matrix is derived.
{"title":"The Solutions of Matrix Equations (AX = B, XC = D) with a Submatrix Constraint","authors":"Fan-Liang Li, Xiyan Hu, Lei Zhang","doi":"10.1109/ICNC.2008.535","DOIUrl":"https://doi.org/10.1109/ICNC.2008.535","url":null,"abstract":"In this paper, the solvability conditions and the general solution of matrix equations (AX=B,XC=D) with a submatrix constraint are obtained by using the SVD(singular value decomposition) of matrix. In addition, the expression of the optimal approximation solution to a given matrix is derived.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"20 1","pages":"381-383"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89846775","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}
SVM is a novel statistical learning method that has been successfully applied in speaker recognition. However, Extractive feature vectors from the speech are overlapped and noisy is included in the original data space, these problems can lead to experience difficulties, training complication during training SVM, and the result will be reduced during the recognition phase. In this paper, a novel method is proposed to reduce the noise and input vectors of the SVM. Firstly data dimensions are reduced and noise is removed by using PCA transform, secondly feature data are selected at boundary of each cluster as SVs by using Kernel-based fuzzy clustering technique. The training data, time and storage can be reduced remarkably compared with traditional SVM; the speaker identification system based on our proposed reduced support vector machine (RSVM) has better robustness compared with other reduced algorithms.
{"title":"A Novel Reduction Method for Text-Independent Speaker Identification","authors":"Yan Wang, Xue Liu, Yujuan Xing, Ming Li","doi":"10.1109/ICNC.2008.708","DOIUrl":"https://doi.org/10.1109/ICNC.2008.708","url":null,"abstract":"SVM is a novel statistical learning method that has been successfully applied in speaker recognition. However, Extractive feature vectors from the speech are overlapped and noisy is included in the original data space, these problems can lead to experience difficulties, training complication during training SVM, and the result will be reduced during the recognition phase. In this paper, a novel method is proposed to reduce the noise and input vectors of the SVM. Firstly data dimensions are reduced and noise is removed by using PCA transform, secondly feature data are selected at boundary of each cluster as SVs by using Kernel-based fuzzy clustering technique. The training data, time and storage can be reduced remarkably compared with traditional SVM; the speaker identification system based on our proposed reduced support vector machine (RSVM) has better robustness compared with other reduced algorithms.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"31 1","pages":"66-70"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89932807","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}
Yunong Zhang, Ke Chen, Xuezhong Li, Chengfu Yi, Hong Zhu
In view of the great potential in parallel processing and ready implementation via hardware, neural networks are now often employed to solve online matrix algebraic problems. Recently, a special kind of recurrent neural network has been proposed by Zhang et al, which could be generalized to solving online Lyapunov equation with time-varying coefficient matrices. In comparison with gradient-based neural networks (GNN), the resultant Zhang neural networks (ZNN) perform much better on solving these time-varying problems. This paper investigates the MATLAB Simulink modeling, simulative verification and comparison of ZNN and GNN models for time-varying Lyapunov equation solving. Computer-simulation results verify that superior convergence and efficacy could be achieved by such ZNN models when solving the time-varying Lyapunov matrix equation, as compared to the GNN models.
{"title":"Simulink Modeling and Comparison of Zhang Neural Networks and Gradient Neural Networks for Time-Varying Lyapunov Equation Solving","authors":"Yunong Zhang, Ke Chen, Xuezhong Li, Chengfu Yi, Hong Zhu","doi":"10.1109/ICNC.2008.47","DOIUrl":"https://doi.org/10.1109/ICNC.2008.47","url":null,"abstract":"In view of the great potential in parallel processing and ready implementation via hardware, neural networks are now often employed to solve online matrix algebraic problems. Recently, a special kind of recurrent neural network has been proposed by Zhang et al, which could be generalized to solving online Lyapunov equation with time-varying coefficient matrices. In comparison with gradient-based neural networks (GNN), the resultant Zhang neural networks (ZNN) perform much better on solving these time-varying problems. This paper investigates the MATLAB Simulink modeling, simulative verification and comparison of ZNN and GNN models for time-varying Lyapunov equation solving. Computer-simulation results verify that superior convergence and efficacy could be achieved by such ZNN models when solving the time-varying Lyapunov matrix equation, as compared to the GNN models.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"8 1","pages":"521-525"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89973042","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}
Estimating landmarks corresponding plays a key role in landmark-based multimodal image registration. In this paper, a novel landmarks corresponding estimation in multimodal image registration using mean shift algorithm is proposed. Edge feature potential is defined to transform images from intensity feature space to edge structure feature space. Image corner points are detected as candidate landmarks. Mean shift iterations are adopted to search the most probable corresponding point positions in the two images based on the edge structure feature. Moreover, mutual information between two local regions is computed to eliminate mis-matching landmarks. Finally, the source images are transformed by compact support thin-plate spline interpolation. Experiments show that the precision in location of corresponding landmarks is satisfied. The proposed technique is feasible and rapid shown in the experiments of various multi-modal medical images registration.
{"title":"Multimodal Medical Image Elastic Registration Using Mean Shift","authors":"Xuan S. Yang, J. Pei","doi":"10.1109/ICNC.2008.159","DOIUrl":"https://doi.org/10.1109/ICNC.2008.159","url":null,"abstract":"Estimating landmarks corresponding plays a key role in landmark-based multimodal image registration. In this paper, a novel landmarks corresponding estimation in multimodal image registration using mean shift algorithm is proposed. Edge feature potential is defined to transform images from intensity feature space to edge structure feature space. Image corner points are detected as candidate landmarks. Mean shift iterations are adopted to search the most probable corresponding point positions in the two images based on the edge structure feature. Moreover, mutual information between two local regions is computed to eliminate mis-matching landmarks. Finally, the source images are transformed by compact support thin-plate spline interpolation. Experiments show that the precision in location of corresponding landmarks is satisfied. The proposed technique is feasible and rapid shown in the experiments of various multi-modal medical images registration.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"1 1","pages":"177-181"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90835343","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：