Pub Date : 2002-11-18DOI: 10.1109/ICONIP.2002.1202826
C. Leung, Sen Jiang
We proposed a joint source-channel coded image-transmission scheme over an inter-symbol interference (ISI) channel. The scheme uses a self-organized-map (SOM) to encode the source data and a convolutional code to protect the quantized data. So, we can regard the ISI channel as another convolutional code. The combination of convolutional code, interleaver and ISI channel equals to a serial turbo code. In the receiver side, we decode the serial turbo code iteratively. The soft-output of the turbo decoder is utilized to get better-reconstructed images by soft decision. In our approach, when a new codebook is used or the noise power of the channels changes, we do not need to carry out the association between codevectors and transmission binary words. By performance comparison, it proves our scheme is comparable to the traditional channel-optimized vector quantization (COVQ) scheme.
{"title":"Soft-decoding for self-organized map","authors":"C. Leung, Sen Jiang","doi":"10.1109/ICONIP.2002.1202826","DOIUrl":"https://doi.org/10.1109/ICONIP.2002.1202826","url":null,"abstract":"We proposed a joint source-channel coded image-transmission scheme over an inter-symbol interference (ISI) channel. The scheme uses a self-organized-map (SOM) to encode the source data and a convolutional code to protect the quantized data. So, we can regard the ISI channel as another convolutional code. The combination of convolutional code, interleaver and ISI channel equals to a serial turbo code. In the receiver side, we decode the serial turbo code iteratively. The soft-output of the turbo decoder is utilized to get better-reconstructed images by soft decision. In our approach, when a new codebook is used or the noise power of the channels changes, we do not need to carry out the association between codevectors and transmission binary words. By performance comparison, it proves our scheme is comparable to the traditional channel-optimized vector quantization (COVQ) scheme.","PeriodicalId":146553,"journal":{"name":"Proceedings of the 9th International Conference on Neural Information Processing, 2002. ICONIP '02.","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128342945","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 : 2002-11-18DOI: 10.1109/ICONIP.2002.1199052
Leo Chin Sim, H. Schroder, G. Leedham
This paper describes an implementation of a novel line removal Hough transform on a new parallel architectural system, referred to as a hybrid system. The algorithm, which is used for detecting lines in an image, strips away major line so that minor lines can be more easily detectable. The algorithm was subsequently implemented on the hybrid system. A hybrid system is a combination of SIMD and MIMD system processing the data in parallel. In this paper, we also introduce a new SIMD concept.
{"title":"Fast line detection using major line removal morphological Hough transform","authors":"Leo Chin Sim, H. Schroder, G. Leedham","doi":"10.1109/ICONIP.2002.1199052","DOIUrl":"https://doi.org/10.1109/ICONIP.2002.1199052","url":null,"abstract":"This paper describes an implementation of a novel line removal Hough transform on a new parallel architectural system, referred to as a hybrid system. The algorithm, which is used for detecting lines in an image, strips away major line so that minor lines can be more easily detectable. The algorithm was subsequently implemented on the hybrid system. A hybrid system is a combination of SIMD and MIMD system processing the data in parallel. In this paper, we also introduce a new SIMD concept.","PeriodicalId":146553,"journal":{"name":"Proceedings of the 9th International Conference on Neural Information Processing, 2002. ICONIP '02.","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126898572","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 : 2002-11-18DOI: 10.1109/ICONIP.2002.1202823
Cong Wang, Guanrong Chen, S. Ge, D. Hill
In this paper, by combining smart neural design with a recently proposed ISS-modular neural control approach, we present a smart neural control scheme for general (non-affine) pure-feedback systems. Although the neural controller in achieves a semi-global result for general (non-affine) pure-feedback systems, it is by nature a high-order dynamic controller, which cannot be reduced in general due to its need of simultaneous adaptation of a large number of neural weights. To overcome this problem, in this paper we develop a smart neural controller, which on the contrary is a static and low-order controller, hence more computationally feasible in practical design and implementation. To improve the NN generalization ability, which plays an important role in our smart neural control scheme, chaotic reference signals are employed in the training phase of the scheme, where the complex chaotic signals offer much richer information for NN learning due to the ergodicity of chaos. Since pure-feedback system represents a very large class of nonlinear systems, the smart neural control scheme is expected to be useful for a wide variety of industrial applications.
{"title":"Smart neural control of pure-feedback systems","authors":"Cong Wang, Guanrong Chen, S. Ge, D. Hill","doi":"10.1109/ICONIP.2002.1202823","DOIUrl":"https://doi.org/10.1109/ICONIP.2002.1202823","url":null,"abstract":"In this paper, by combining smart neural design with a recently proposed ISS-modular neural control approach, we present a smart neural control scheme for general (non-affine) pure-feedback systems. Although the neural controller in achieves a semi-global result for general (non-affine) pure-feedback systems, it is by nature a high-order dynamic controller, which cannot be reduced in general due to its need of simultaneous adaptation of a large number of neural weights. To overcome this problem, in this paper we develop a smart neural controller, which on the contrary is a static and low-order controller, hence more computationally feasible in practical design and implementation. To improve the NN generalization ability, which plays an important role in our smart neural control scheme, chaotic reference signals are employed in the training phase of the scheme, where the complex chaotic signals offer much richer information for NN learning due to the ergodicity of chaos. Since pure-feedback system represents a very large class of nonlinear systems, the smart neural control scheme is expected to be useful for a wide variety of industrial applications.","PeriodicalId":146553,"journal":{"name":"Proceedings of the 9th International Conference on Neural Information Processing, 2002. ICONIP '02.","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129285202","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 : 2002-11-18DOI: 10.1109/ICONIP.2002.1202178
N. Sato, Y. Yamaguchi
We showed that phase coding of behavioral temporal sequences in the form of theta phase precession has high advantage for the memory storage in the hippocampus. In this paper the neural dynamics of phase coding is applied to object-place memory, a typical hippocampus dependent memory in human. We assume that visual input consists of central and peripheral structures in visual field which encode object and scene respectively. The temporal sequence these inputs generated by saccadic eye movement is sent to our hippocampal network. Results of our computer experiments demonstrate that object-place memory can be instantaneously stored by phase precession dynamics. It is suggested that the phase precession in rat hippocampus could be generalized as a fundamental mechanism in human hippocampal memory.
{"title":"A neural network model of the hippocampus with theta phase precession for object-place memory","authors":"N. Sato, Y. Yamaguchi","doi":"10.1109/ICONIP.2002.1202178","DOIUrl":"https://doi.org/10.1109/ICONIP.2002.1202178","url":null,"abstract":"We showed that phase coding of behavioral temporal sequences in the form of theta phase precession has high advantage for the memory storage in the hippocampus. In this paper the neural dynamics of phase coding is applied to object-place memory, a typical hippocampus dependent memory in human. We assume that visual input consists of central and peripheral structures in visual field which encode object and scene respectively. The temporal sequence these inputs generated by saccadic eye movement is sent to our hippocampal network. Results of our computer experiments demonstrate that object-place memory can be instantaneously stored by phase precession dynamics. It is suggested that the phase precession in rat hippocampus could be generalized as a fundamental mechanism in human hippocampal memory.","PeriodicalId":146553,"journal":{"name":"Proceedings of the 9th International Conference on Neural Information Processing, 2002. ICONIP '02.","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114346572","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 : 2002-11-18DOI: 10.1109/ICONIP.2002.1202890
G. Ashida, M. Kubo
A new neuronal model based on the stochastic activity of a single channel is proposed. The model connects the microscopic ionic kinetics to the macroscopic neuronal behavior and describes a neuron as a organized structure of individual subneuronal elements. This illustration has not been well achieved by the common models employed so far. Simulational results by using this novel model shows that spatially distributed ion channels can reproduce the neuronal dynamics - the existence of the stable state, initiation and propagation of the action potentials. The membrane potential measured at different points enables us to estimate the propagation speed of the action potentials.
{"title":"Neuronal dynamics based on individual stochastic ion channels","authors":"G. Ashida, M. Kubo","doi":"10.1109/ICONIP.2002.1202890","DOIUrl":"https://doi.org/10.1109/ICONIP.2002.1202890","url":null,"abstract":"A new neuronal model based on the stochastic activity of a single channel is proposed. The model connects the microscopic ionic kinetics to the macroscopic neuronal behavior and describes a neuron as a organized structure of individual subneuronal elements. This illustration has not been well achieved by the common models employed so far. Simulational results by using this novel model shows that spatially distributed ion channels can reproduce the neuronal dynamics - the existence of the stable state, initiation and propagation of the action potentials. The membrane potential measured at different points enables us to estimate the propagation speed of the action potentials.","PeriodicalId":146553,"journal":{"name":"Proceedings of the 9th International Conference on Neural Information Processing, 2002. ICONIP '02.","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114364307","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 : 2002-11-18DOI: 10.1109/ICONIP.2002.1202839
P. Luo, K. Wong
We analyse the dynamics of on-line learning with recycled examples using the cavity method. For large input dimensions, we derive equations for the macroscopic parameters, providing solutions to Adaline learning as a benchmark. Comparison with batch learning is discussed.
{"title":"Online learning with recycled examples: a cavity analysis","authors":"P. Luo, K. Wong","doi":"10.1109/ICONIP.2002.1202839","DOIUrl":"https://doi.org/10.1109/ICONIP.2002.1202839","url":null,"abstract":"We analyse the dynamics of on-line learning with recycled examples using the cavity method. For large input dimensions, we derive equations for the macroscopic parameters, providing solutions to Adaline learning as a benchmark. Comparison with batch learning is discussed.","PeriodicalId":146553,"journal":{"name":"Proceedings of the 9th International Conference on Neural Information Processing, 2002. ICONIP '02.","volume":"211 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116295134","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 : 2002-11-18DOI: 10.1109/ICONIP.2002.1198965
Xieping Gao
Based on the wavelet and multiwavelet theory, some new notions of wavelet neural networks and multiwavelet neural networks have been proposed as an alternative to feedforward neural networks for approximating arbitrary nonlinear functions by some scholars recently. In this paper we give a comparative research on (multi-)wavelet neural networks,especially emphasized on their mathematical basis,the systematic synthesis, and the efficacy analysis,such as the "curse of dimensionality". Some experiments are given to verify the efficiency of wavelet neural networks.
{"title":"A comparative research on wavelet neural networks","authors":"Xieping Gao","doi":"10.1109/ICONIP.2002.1198965","DOIUrl":"https://doi.org/10.1109/ICONIP.2002.1198965","url":null,"abstract":"Based on the wavelet and multiwavelet theory, some new notions of wavelet neural networks and multiwavelet neural networks have been proposed as an alternative to feedforward neural networks for approximating arbitrary nonlinear functions by some scholars recently. In this paper we give a comparative research on (multi-)wavelet neural networks,especially emphasized on their mathematical basis,the systematic synthesis, and the efficacy analysis,such as the \"curse of dimensionality\". Some experiments are given to verify the efficiency of wavelet neural networks.","PeriodicalId":146553,"journal":{"name":"Proceedings of the 9th International Conference on Neural Information Processing, 2002. ICONIP '02.","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116341279","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 : 2002-11-18DOI: 10.1109/ICONIP.2002.1202824
Jinglu Hu, K. Hirasawa
This paper introduces a new method for applying multilayer perceptron (MLP) network to control of nonlinear systems. The MLP network is not used directly as a nonlinear controller, but used indirectly via an ARX-like macro-model. The ARX-like model incorporating MLP network is constructed in such a way that it has similar linear properties to a linear ARX model. The nonlinear controller is then designed in the same way as designing a linear controller based on a linear ARX model. Numerical simulations are carried to demonstrate the effectiveness of the new method.
{"title":"A method for applying multilayer perceptrons to control of nonlinear systems","authors":"Jinglu Hu, K. Hirasawa","doi":"10.1109/ICONIP.2002.1202824","DOIUrl":"https://doi.org/10.1109/ICONIP.2002.1202824","url":null,"abstract":"This paper introduces a new method for applying multilayer perceptron (MLP) network to control of nonlinear systems. The MLP network is not used directly as a nonlinear controller, but used indirectly via an ARX-like macro-model. The ARX-like model incorporating MLP network is constructed in such a way that it has similar linear properties to a linear ARX model. The nonlinear controller is then designed in the same way as designing a linear controller based on a linear ARX model. Numerical simulations are carried to demonstrate the effectiveness of the new method.","PeriodicalId":146553,"journal":{"name":"Proceedings of the 9th International Conference on Neural Information Processing, 2002. ICONIP '02.","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126332074","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 : 2002-11-18DOI: 10.1109/ICONIP.2002.1198209
K. Chiu, L. Xu
A recently developed factor analytic technique, non-Gaussian factor analysis (NFA) is found to be well-suited for the analysis of classical arbitrage pricing theory (APT). In particular, the model selection ability of NFA is of substantial benefit to the critical task of factor number determination in traditional APT analysis. We aim to demonstrate how a reasonable number of independent factors in APT could be determined by applying the NFA technique. Experimental comparisons with the other two conventional methods are illustrated to confirm the superiority of this new approach.
{"title":"Number of statistical independent factors in arbitrage pricing theory from the perspective of non-Gaussian factor analysis","authors":"K. Chiu, L. Xu","doi":"10.1109/ICONIP.2002.1198209","DOIUrl":"https://doi.org/10.1109/ICONIP.2002.1198209","url":null,"abstract":"A recently developed factor analytic technique, non-Gaussian factor analysis (NFA) is found to be well-suited for the analysis of classical arbitrage pricing theory (APT). In particular, the model selection ability of NFA is of substantial benefit to the critical task of factor number determination in traditional APT analysis. We aim to demonstrate how a reasonable number of independent factors in APT could be determined by applying the NFA technique. Experimental comparisons with the other two conventional methods are illustrated to confirm the superiority of this new approach.","PeriodicalId":146553,"journal":{"name":"Proceedings of the 9th International Conference on Neural Information Processing, 2002. ICONIP '02.","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127962063","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 : 2002-11-18DOI: 10.1109/ICONIP.2002.1198987
A. Paplinski, L. Gustafsson
Autism is a developmental disorder in which attention shift impairment and strong familiarity preference are considered to be prime deficiencies. We model these two characteristics of autistic behaviour using Self-Organizing Maps (SOFM).
{"title":"An attempt in modelling autism using self-organizing maps","authors":"A. Paplinski, L. Gustafsson","doi":"10.1109/ICONIP.2002.1198987","DOIUrl":"https://doi.org/10.1109/ICONIP.2002.1198987","url":null,"abstract":"Autism is a developmental disorder in which attention shift impairment and strong familiarity preference are considered to be prime deficiencies. We model these two characteristics of autistic behaviour using Self-Organizing Maps (SOFM).","PeriodicalId":146553,"journal":{"name":"Proceedings of the 9th International Conference on Neural Information Processing, 2002. ICONIP '02.","volume":"311 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2002-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115936311","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：