Pub Date : 1997-07-21DOI: 10.1109/HOST.1997.613483
W. Collis, M. Weston, P. White
Linear filter theory based on Wiener filtering is well understood and used widely in many fields of image and signal processing. However, the use of linear filters is generally associated with implicit approximations. Therefore, in this work a series of non-linear filters is developed based on the concepts of Volterra series and these are applied to image interpolation problems. More explicitly the aim is to interpolate one field of a frame of a television picture to form an estimate of the second field. This is known as de-interlacing and is useful in many areas of video processing, for example standards conversion. Conventional de-interlacing systems use a fixed linear combination of the pixels in the aperture. In this paper we consider the extension of these methods to allow estimators based non-linear combinations of pixel values.
{"title":"The application of non-linear Volterra type filters to television images","authors":"W. Collis, M. Weston, P. White","doi":"10.1109/HOST.1997.613483","DOIUrl":"https://doi.org/10.1109/HOST.1997.613483","url":null,"abstract":"Linear filter theory based on Wiener filtering is well understood and used widely in many fields of image and signal processing. However, the use of linear filters is generally associated with implicit approximations. Therefore, in this work a series of non-linear filters is developed based on the concepts of Volterra series and these are applied to image interpolation problems. More explicitly the aim is to interpolate one field of a frame of a television picture to form an estimate of the second field. This is known as de-interlacing and is useful in many areas of video processing, for example standards conversion. Conventional de-interlacing systems use a fixed linear combination of the pixels in the aperture. In this paper we consider the extension of these methods to allow estimators based non-linear combinations of pixel values.","PeriodicalId":305928,"journal":{"name":"Proceedings of the IEEE Signal Processing Workshop on Higher-Order Statistics","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128087994","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 : 1997-07-21DOI: 10.1109/HOST.1997.613545
S. Choi, A. Cichocki
We presents a new necessary and sufficient condition for the blind separation of sources having non-zero kurtosis, from their linear mixtures. It is shown here that a new blind separation criterion based on both odd (f(y)=y/sup 3/) and even (f(y)=y/sup 2/) functions, presents desirable solutions, provided that all source signals have negative kurtosis (sub-Gaussian) or have positive kurtosis (super-Gaussian). Based on this new separation criterion, a linear feedforward network with lateral feedback connections is constructed. Both theoretical and computer simulation results are presented.
{"title":"A linear feedforward neural network with lateral feedback connections for blind source separation","authors":"S. Choi, A. Cichocki","doi":"10.1109/HOST.1997.613545","DOIUrl":"https://doi.org/10.1109/HOST.1997.613545","url":null,"abstract":"We presents a new necessary and sufficient condition for the blind separation of sources having non-zero kurtosis, from their linear mixtures. It is shown here that a new blind separation criterion based on both odd (f(y)=y/sup 3/) and even (f(y)=y/sup 2/) functions, presents desirable solutions, provided that all source signals have negative kurtosis (sub-Gaussian) or have positive kurtosis (super-Gaussian). Based on this new separation criterion, a linear feedforward network with lateral feedback connections is constructed. Both theoretical and computer simulation results are presented.","PeriodicalId":305928,"journal":{"name":"Proceedings of the IEEE Signal Processing Workshop on Higher-Order Statistics","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123455059","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 : 1997-07-21DOI: 10.1109/HOST.1997.613487
M. Ashouri
In this paper, two isolated word recognition methods based on high-order statistics and a time-delay neural network (TDNN) for recognition of Farsi spoken digits have been studied. The adopted speech recognition system consists of four modules, namely, a preprocessor, endpoints' detector, feature extractor and classifier. The first method estimates the AR parameters of speech based on the third- and fourth-order cumulants using high-order Yule-Walker, W-slice and 1-D slice approaches. In the second, method, statistical features are extracted from the estimated high-order probability density function (pdf) of thresholded amplitude features. For each pdf estimate, the values of mean, variance, third order moment and entropy are computed. The total number of features for each frame of approximate length of 15 ms is 16. The adopted TDNN has 16 nodes in its input layer, 10 nodes in its output layer and two hidden layers. The learning rule of the adopted TDNN that is based on the backpropagation rule has been modified to decrease the training time. Computer simulation results obtained from recognizing 10 Farsi digits spoken by different speakers shows that the first method has a better recognition rate while the second method necessitates less computation.
{"title":"Isolated word recognition using high-order statistics and time-delay neural networks","authors":"M. Ashouri","doi":"10.1109/HOST.1997.613487","DOIUrl":"https://doi.org/10.1109/HOST.1997.613487","url":null,"abstract":"In this paper, two isolated word recognition methods based on high-order statistics and a time-delay neural network (TDNN) for recognition of Farsi spoken digits have been studied. The adopted speech recognition system consists of four modules, namely, a preprocessor, endpoints' detector, feature extractor and classifier. The first method estimates the AR parameters of speech based on the third- and fourth-order cumulants using high-order Yule-Walker, W-slice and 1-D slice approaches. In the second, method, statistical features are extracted from the estimated high-order probability density function (pdf) of thresholded amplitude features. For each pdf estimate, the values of mean, variance, third order moment and entropy are computed. The total number of features for each frame of approximate length of 15 ms is 16. The adopted TDNN has 16 nodes in its input layer, 10 nodes in its output layer and two hidden layers. The learning rule of the adopted TDNN that is based on the backpropagation rule has been modified to decrease the training time. Computer simulation results obtained from recognizing 10 Farsi digits spoken by different speakers shows that the first method has a better recognition rate while the second method necessitates less computation.","PeriodicalId":305928,"journal":{"name":"Proceedings of the IEEE Signal Processing Workshop on Higher-Order Statistics","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123708806","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 : 1997-07-21DOI: 10.1109/HOST.1997.613514
B. Porat, B. Friedlander
This paper proposes a scheme for maximum-likelihood-based signal reconstruction. The scheme extends a previous work by Yellin and Friedlander to the case of fractionally-spaced data. The use of fractionally-spaced data obviates the need for timing-phase recovery. Batch and adaptive algorithms are derived and illustrated by examples. The algorithms are useful for equalization of digital communication channels.
{"title":"Fractionally-spaced signal reconstruction based on maximum likelihood","authors":"B. Porat, B. Friedlander","doi":"10.1109/HOST.1997.613514","DOIUrl":"https://doi.org/10.1109/HOST.1997.613514","url":null,"abstract":"This paper proposes a scheme for maximum-likelihood-based signal reconstruction. The scheme extends a previous work by Yellin and Friedlander to the case of fractionally-spaced data. The use of fractionally-spaced data obviates the need for timing-phase recovery. Batch and adaptive algorithms are derived and illustrated by examples. The algorithms are useful for equalization of digital communication channels.","PeriodicalId":305928,"journal":{"name":"Proceedings of the IEEE Signal Processing Workshop on Higher-Order Statistics","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130285734","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 : 1997-07-21DOI: 10.1109/HOST.1997.613497
Ta‐Hsin Li, Key-Shin Lii
A method is proposed for the restoration of linearly blurred two-tone images without requiring the knowledge of the blur parameters. The method jointly estimates the original image and the blur parameters based on some statistical parameters at the output of an inverse filter. Unlike some other blind image restoration procedures, the proposed method does not require the estimation or modeling of the statistical properties of the original image, yet can be justified even for non-i.i.d. images.
{"title":"Deblurring two-tone images by a joint estimation approach using higher-order statistics","authors":"Ta‐Hsin Li, Key-Shin Lii","doi":"10.1109/HOST.1997.613497","DOIUrl":"https://doi.org/10.1109/HOST.1997.613497","url":null,"abstract":"A method is proposed for the restoration of linearly blurred two-tone images without requiring the knowledge of the blur parameters. The method jointly estimates the original image and the blur parameters based on some statistical parameters at the output of an inverse filter. Unlike some other blind image restoration procedures, the proposed method does not require the estimation or modeling of the statistical properties of the original image, yet can be justified even for non-i.i.d. images.","PeriodicalId":305928,"journal":{"name":"Proceedings of the IEEE Signal Processing Workshop on Higher-Order Statistics","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134426295","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 : 1997-07-21DOI: 10.1109/HOST.1997.613512
S. Chen, S. McLaughlin
A family of blind equalisation algorithms identifies a channel model based on a higher-order cumulant (HOC) fitting approach. Since HOC cost functions are multimodal, gradient search techniques require a good initial estimate to avoid converging to local minima. We present a blind identification scheme which uses genetic algorithms (GAs) to optimise a HOC cost function. Because GAs are efficient global optimal search strategies, the proposed method guarantees to find a global optimal channel estimate. A micro-GA implementation is adopted to further enhance computational efficiency. As is demonstrated in computer simulation, this GA based scheme is robust and accurate, and has a fast convergence performance.
{"title":"Blind channel identification based on higher-order cumulant fitting using genetic algorithms","authors":"S. Chen, S. McLaughlin","doi":"10.1109/HOST.1997.613512","DOIUrl":"https://doi.org/10.1109/HOST.1997.613512","url":null,"abstract":"A family of blind equalisation algorithms identifies a channel model based on a higher-order cumulant (HOC) fitting approach. Since HOC cost functions are multimodal, gradient search techniques require a good initial estimate to avoid converging to local minima. We present a blind identification scheme which uses genetic algorithms (GAs) to optimise a HOC cost function. Because GAs are efficient global optimal search strategies, the proposed method guarantees to find a global optimal channel estimate. A micro-GA implementation is adopted to further enhance computational efficiency. As is demonstrated in computer simulation, this GA based scheme is robust and accurate, and has a fast convergence performance.","PeriodicalId":305928,"journal":{"name":"Proceedings of the IEEE Signal Processing Workshop on Higher-Order Statistics","volume":"77 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134004322","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 : 1997-07-21DOI: 10.1109/HOST.1997.613554
Chong-Yung Chi
In the paper, a parametric Fourier series based model (FSBM) for or as an approximation to an arbitrary nonminimum-phase linear time-invariant (LTI) system is proposed for statistical signal processing applications where a model for LTI systems is needed. Based on the FSBM, a (minimum-phase) linear prediction error (LPE) filter for amplitude estimation of the unknown LTI system together with the Cramer Rao (CR) bounds is presented. Then an iterative algorithm for obtaining the optimum mean-square LPE filter with finite data is presented which is also an approximate maximum likelihood algorithm when the data are Gaussian. Then three iterative algorithms using higher-order statistics with finite non-Gaussian data are presented for estimating parameters of the FSBM followed by some simulation results to support the efficacy of the proposed algorithms. Finally, we draw some conclusions.
{"title":"Fourier series based nonminimum phase model for second- and higher-order statistical signal processing","authors":"Chong-Yung Chi","doi":"10.1109/HOST.1997.613554","DOIUrl":"https://doi.org/10.1109/HOST.1997.613554","url":null,"abstract":"In the paper, a parametric Fourier series based model (FSBM) for or as an approximation to an arbitrary nonminimum-phase linear time-invariant (LTI) system is proposed for statistical signal processing applications where a model for LTI systems is needed. Based on the FSBM, a (minimum-phase) linear prediction error (LPE) filter for amplitude estimation of the unknown LTI system together with the Cramer Rao (CR) bounds is presented. Then an iterative algorithm for obtaining the optimum mean-square LPE filter with finite data is presented which is also an approximate maximum likelihood algorithm when the data are Gaussian. Then three iterative algorithms using higher-order statistics with finite non-Gaussian data are presented for estimating parameters of the FSBM followed by some simulation results to support the efficacy of the proposed algorithms. Finally, we draw some conclusions.","PeriodicalId":305928,"journal":{"name":"Proceedings of the IEEE Signal Processing Workshop on Higher-Order Statistics","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133276145","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 : 1997-07-21DOI: 10.1109/HOST.1997.613516
Y. Inoue, T. Sato
Blind deconvolution and blind equalization have been important interesting topics in diverse fields including data communication, image processing and geophysical data processing. Inouye and Habe (1995) proposed a constrained multistage criterion for attaining blind deconvolution of multichannel linear time-invariant (LTI) systems. In this paper, based on their constrained criterion, we present an iterative algorithm for solving the blind deconvolution problem of multichannel LTI systems. Inouye and Sato (1996) proposed new unconstrained criteria for accomplishing the blind deconvolution of multichannel LTI systems. Based on their unconstrained criteria, we show iterative algorithms for solving the blind deconvolution of multichannel LTI systems. Simulation examples are included to examine the proposed algorithms.
{"title":"On-line algorithms for blind deconvolution of multichannel linear time-invariant systems","authors":"Y. Inoue, T. Sato","doi":"10.1109/HOST.1997.613516","DOIUrl":"https://doi.org/10.1109/HOST.1997.613516","url":null,"abstract":"Blind deconvolution and blind equalization have been important interesting topics in diverse fields including data communication, image processing and geophysical data processing. Inouye and Habe (1995) proposed a constrained multistage criterion for attaining blind deconvolution of multichannel linear time-invariant (LTI) systems. In this paper, based on their constrained criterion, we present an iterative algorithm for solving the blind deconvolution problem of multichannel LTI systems. Inouye and Sato (1996) proposed new unconstrained criteria for accomplishing the blind deconvolution of multichannel LTI systems. Based on their unconstrained criteria, we show iterative algorithms for solving the blind deconvolution of multichannel LTI systems. Simulation examples are included to examine the proposed algorithms.","PeriodicalId":305928,"journal":{"name":"Proceedings of the IEEE Signal Processing Workshop on Higher-Order Statistics","volume":"159 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116506494","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 : 1997-07-21DOI: 10.1109/HOST.1997.613499
M. A. Shcherbakov
A method for identification of discrete nonlinear systems in terms of the Volterra-Wiener series is presented. It is shown that use of a special composite-frequency input signal as an approximation to Gaussian noise provides the computational efficiency of this method especially for high order kernels. Orthogonal functionals and consistent estimates for Wiener kernels in the frequency domain are derived for this class of noise input. The basis of the proposed computational procedure for practical identification is the fast Fourier transform (FFT) algorithm which is used both for generation of actions and for analysis of system reactions.
{"title":"Fast estimation of Wiener kernels of nonlinear systems in the frequency domain","authors":"M. A. Shcherbakov","doi":"10.1109/HOST.1997.613499","DOIUrl":"https://doi.org/10.1109/HOST.1997.613499","url":null,"abstract":"A method for identification of discrete nonlinear systems in terms of the Volterra-Wiener series is presented. It is shown that use of a special composite-frequency input signal as an approximation to Gaussian noise provides the computational efficiency of this method especially for high order kernels. Orthogonal functionals and consistent estimates for Wiener kernels in the frequency domain are derived for this class of noise input. The basis of the proposed computational procedure for practical identification is the fast Fourier transform (FFT) algorithm which is used both for generation of actions and for analysis of system reactions.","PeriodicalId":305928,"journal":{"name":"Proceedings of the IEEE Signal Processing Workshop on Higher-Order Statistics","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114768870","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 : 1997-07-21DOI: 10.1109/HOST.1997.613555
L. A. Pflug, G. Ioup, J. Ioup, P. Jackson
Many underwater acoustic signal processing algorithms are designed for use in stationary and/or Gaussian noise. While these assumptions are often valid for applications in deep water ocean areas, they may not be appropriate for shallow water areas, especially in the presence of local shipping activity. Local shipping also produces spatial correlation in the noise and introduces additional complexity for multichannel processing. In this paper, two 30-minute sets of ambient ocean noise, recorded near the San Diego, California coast, are analyzed for stationarity and Gaussianity using the Kolmogorov-Smirnov test. Since processing algorithms based on higher order statistics often assume Gaussianity, time-dependent fluctuations in the third and fourth order cumulants are also analyzed. The analysis reveals significant variability in the time lengths of stationary periods, and episodic periods of nonGaussianity that last for up to five minutes. Statistical fluctuations appear predominantly in the second and fourth order cumulants rather than the third order cumulant. The shipping noise is also shown to be correlated between pairs of hydrophones with the level of correlation varying over time and the correlation ranging from positive to negative with increasing channel separation.
{"title":"Variability in higher order statistics of measured shallow-water shipping noise","authors":"L. A. Pflug, G. Ioup, J. Ioup, P. Jackson","doi":"10.1109/HOST.1997.613555","DOIUrl":"https://doi.org/10.1109/HOST.1997.613555","url":null,"abstract":"Many underwater acoustic signal processing algorithms are designed for use in stationary and/or Gaussian noise. While these assumptions are often valid for applications in deep water ocean areas, they may not be appropriate for shallow water areas, especially in the presence of local shipping activity. Local shipping also produces spatial correlation in the noise and introduces additional complexity for multichannel processing. In this paper, two 30-minute sets of ambient ocean noise, recorded near the San Diego, California coast, are analyzed for stationarity and Gaussianity using the Kolmogorov-Smirnov test. Since processing algorithms based on higher order statistics often assume Gaussianity, time-dependent fluctuations in the third and fourth order cumulants are also analyzed. The analysis reveals significant variability in the time lengths of stationary periods, and episodic periods of nonGaussianity that last for up to five minutes. Statistical fluctuations appear predominantly in the second and fourth order cumulants rather than the third order cumulant. The shipping noise is also shown to be correlated between pairs of hydrophones with the level of correlation varying over time and the correlation ranging from positive to negative with increasing channel separation.","PeriodicalId":305928,"journal":{"name":"Proceedings of the IEEE Signal Processing Workshop on Higher-Order Statistics","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134572835","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：