Pub Date : 1994-04-19DOI: 10.1109/ICASSP.1994.389351
A. Samouelian
This paper presents a knowledge based approach to consonant recognition. In traditional knowledge based systems, the expert is the linguist/phonetician who attempts to describe and quantify the acoustic events, in the form of production rules into phonetic description. This paper proposes to alter the expert's role so that the expert only needs to provide the basic structure of the phonetic classification. The knowledge itself can then be induced from examples in the agreed structure. Thus the acoustic-phonetic rules are moved from the expert's head to the machine memory via the language of examples rather than via the language of explicit articulation. Recognition results on three broad phonetic classes, namely plosives, semi-vowels and nasals, for a combination of feature sets, for speaker dependent and independent recognition, are presented.<>
{"title":"Knowledge based approach to consonant recognition","authors":"A. Samouelian","doi":"10.1109/ICASSP.1994.389351","DOIUrl":"https://doi.org/10.1109/ICASSP.1994.389351","url":null,"abstract":"This paper presents a knowledge based approach to consonant recognition. In traditional knowledge based systems, the expert is the linguist/phonetician who attempts to describe and quantify the acoustic events, in the form of production rules into phonetic description. This paper proposes to alter the expert's role so that the expert only needs to provide the basic structure of the phonetic classification. The knowledge itself can then be induced from examples in the agreed structure. Thus the acoustic-phonetic rules are moved from the expert's head to the machine memory via the language of examples rather than via the language of explicit articulation. Recognition results on three broad phonetic classes, namely plosives, semi-vowels and nasals, for a combination of feature sets, for speaker dependent and independent recognition, are presented.<<ETX>>","PeriodicalId":290798,"journal":{"name":"Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127228110","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 : 1994-04-19DOI: 10.1109/ICASSP.1994.389907
Edward A. Lee
In the Fall of 1993 at Berkeley we offered an experimental graduate course that focused on languages for modeling and design of signal processing systems. A major motivation for the course is our Ptolemy project, in which we are experimenting with models of computation and design methodology for signal processing systems. The applicable theory of computation primarily concerns stream datatypes and their implementation in dataflow, functional, and concurrent imperative languages. The issues addressed in the course include determinacy, concurrency, strictness, parallel scheduling, polymorphism, recursion, higher-order functions, and visual syntax. The emphasis is on studying strengths and weaknesses of existing and proposed design environments for signal processing.<>
{"title":"Computing and signal processing: an experimental multidisciplinary course","authors":"Edward A. Lee","doi":"10.1109/ICASSP.1994.389907","DOIUrl":"https://doi.org/10.1109/ICASSP.1994.389907","url":null,"abstract":"In the Fall of 1993 at Berkeley we offered an experimental graduate course that focused on languages for modeling and design of signal processing systems. A major motivation for the course is our Ptolemy project, in which we are experimenting with models of computation and design methodology for signal processing systems. The applicable theory of computation primarily concerns stream datatypes and their implementation in dataflow, functional, and concurrent imperative languages. The issues addressed in the course include determinacy, concurrency, strictness, parallel scheduling, polymorphism, recursion, higher-order functions, and visual syntax. The emphasis is on studying strengths and weaknesses of existing and proposed design environments for signal processing.<<ETX>>","PeriodicalId":290798,"journal":{"name":"Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing","volume":"82 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124879652","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 : 1994-04-19DOI: 10.1109/ICASSP.1994.390020
A. Gottlieb
Asymmetric digital subscriber line (ADSL) is an emerging technology for the transport of asymmetric DS1-rate services over a single copper twisted-pair in the local loop plant. One of the key ingredients to the ADSL system is the upstream low bit-rate control channel. The architecture and laboratory implementation of the low bit-rate upstream control channel is described. The system was first computer modeled, and then the transmitter/receiver were implemented around programmable digital signal processing chips (DSP) such as the TI TMS320C40. Experimental results indicate that the system is functional to loops of 18 kft.<>
{"title":"A DSP-based research prototype reverse channel transmitter/receiver for ADSL","authors":"A. Gottlieb","doi":"10.1109/ICASSP.1994.390020","DOIUrl":"https://doi.org/10.1109/ICASSP.1994.390020","url":null,"abstract":"Asymmetric digital subscriber line (ADSL) is an emerging technology for the transport of asymmetric DS1-rate services over a single copper twisted-pair in the local loop plant. One of the key ingredients to the ADSL system is the upstream low bit-rate control channel. The architecture and laboratory implementation of the low bit-rate upstream control channel is described. The system was first computer modeled, and then the transmitter/receiver were implemented around programmable digital signal processing chips (DSP) such as the TI TMS320C40. Experimental results indicate that the system is functional to loops of 18 kft.<<ETX>>","PeriodicalId":290798,"journal":{"name":"Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126067870","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 : 1994-04-19DOI: 10.1109/ICASSP.1994.389618
K. Konstantinides, B. Natarajan
We present a novel non-linear filtering technique for random noise, based on the principle that random noise is hard to compress. In contrast to spectral filters, the proposed technique does not require a priori knowledge of the noise and signal characteristics. We demonstrate the technique by filtering additive random noise using a compression algorithm based on piecewise linear approximation. A single chip design of the filtering algorithm is also presented. It includes two multiplier-accumulator units, an adder, registers, and a short look-up table. The proposed implementation allows an output sample to be generated every four cycles on average.<>
{"title":"Algorithm and architecture: for non-linear noise filtering via piecewise linear compression","authors":"K. Konstantinides, B. Natarajan","doi":"10.1109/ICASSP.1994.389618","DOIUrl":"https://doi.org/10.1109/ICASSP.1994.389618","url":null,"abstract":"We present a novel non-linear filtering technique for random noise, based on the principle that random noise is hard to compress. In contrast to spectral filters, the proposed technique does not require a priori knowledge of the noise and signal characteristics. We demonstrate the technique by filtering additive random noise using a compression algorithm based on piecewise linear approximation. A single chip design of the filtering algorithm is also presented. It includes two multiplier-accumulator units, an adder, registers, and a short look-up table. The proposed implementation allows an output sample to be generated every four cycles on average.<<ETX>>","PeriodicalId":290798,"journal":{"name":"Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125498723","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 : 1994-04-19DOI: 10.1109/ICASSP.1994.390101
P. Rieder, J. Götze, J. Nossek
An algebraic approach to the design of different kinds of discrete wavelet transforms (orthogonal and biorthogonal single-/multiwavelet transforms, multiwavelet-like transforms) is taken. The different transforms are analysed with respect to computational efforts, approximation properties and symmetry. The design of the orthogonal and biorthogonal single-/multiwavelets requires the solution of a system of linear and nonlinear equations. Only the biorthogonal case enables symmetric coefficients. The basis matrix of the multiwavelet-like transform is easy to compute, orthogonal and ultimately symmetric. Modifications of this multiwavelet-like transform are given with respect to practical applications.<>
{"title":"Algebraic design of discrete multiwavelet transforms","authors":"P. Rieder, J. Götze, J. Nossek","doi":"10.1109/ICASSP.1994.390101","DOIUrl":"https://doi.org/10.1109/ICASSP.1994.390101","url":null,"abstract":"An algebraic approach to the design of different kinds of discrete wavelet transforms (orthogonal and biorthogonal single-/multiwavelet transforms, multiwavelet-like transforms) is taken. The different transforms are analysed with respect to computational efforts, approximation properties and symmetry. The design of the orthogonal and biorthogonal single-/multiwavelets requires the solution of a system of linear and nonlinear equations. Only the biorthogonal case enables symmetric coefficients. The basis matrix of the multiwavelet-like transform is easy to compute, orthogonal and ultimately symmetric. Modifications of this multiwavelet-like transform are given with respect to practical applications.<<ETX>>","PeriodicalId":290798,"journal":{"name":"Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing","volume":"90 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115083214","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 : 1994-04-19DOI: 10.1109/ICASSP.1994.389825
Jiankan Yang, S. Daas, A. Swindlehurst
Blind adaptive algorithms extract signals that overlap in time and frequency by exploiting their temporal structure, but ignore any available spatial (array response) data. On the other hand, direction-finding based methods compute the signal copy weights using estimates of the signal directions, but ignore information about signal structure. In this paper, we present two simple iterative techniques that attempt to incorporate both temporal and spatial information in estimating the signal waveforms received by an array of sensors. The first technique assumes an initial blind signal estimate is available, and uses least-squares to approximate the array response and refine the signal estimate. The second method is applicable to digitally modulated signals, and uses bit decisions made on an initial signal estimate to recompute the signal copy weight vectors. A theoretical performance analysis of both algorithms is conducted for the high SNR case, and some representative simulation results are included.<>
{"title":"Improved signal copy with partially known or unknown array response","authors":"Jiankan Yang, S. Daas, A. Swindlehurst","doi":"10.1109/ICASSP.1994.389825","DOIUrl":"https://doi.org/10.1109/ICASSP.1994.389825","url":null,"abstract":"Blind adaptive algorithms extract signals that overlap in time and frequency by exploiting their temporal structure, but ignore any available spatial (array response) data. On the other hand, direction-finding based methods compute the signal copy weights using estimates of the signal directions, but ignore information about signal structure. In this paper, we present two simple iterative techniques that attempt to incorporate both temporal and spatial information in estimating the signal waveforms received by an array of sensors. The first technique assumes an initial blind signal estimate is available, and uses least-squares to approximate the array response and refine the signal estimate. The second method is applicable to digitally modulated signals, and uses bit decisions made on an initial signal estimate to recompute the signal copy weight vectors. A theoretical performance analysis of both algorithms is conducted for the high SNR case, and some representative simulation results are included.<<ETX>>","PeriodicalId":290798,"journal":{"name":"Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115088732","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 : 1994-04-19DOI: 10.1109/ICASSP.1994.389932
F. Ghorbel, C. Banga
We present the bootstrap sampling techniques applied to some pattern recognition algorithms. Two important procedures in image analysis are tested: a statistical segmentation based on expectation-maximisation (EM) family algorithms and two methods of invariant features extraction for gray level images. In the first case, the results we obtain show that the bootstrap sample selection method gives better results than the classical one both in the quality of the segmented image and the computing time. In the second case, the computation of the moment invariants (MI) and the analytical Fourier Mellin transform (AFMT) by the bootstrap approach using the Monte Carlo approximations are implemented. We note that this approach gives a stable approximation and reduces considerably the computing time, since we select only a small representative sample from the image. These algorithms are applied to natural image (medical image).<>
{"title":"Bootstrap sampling applied to image analysis","authors":"F. Ghorbel, C. Banga","doi":"10.1109/ICASSP.1994.389932","DOIUrl":"https://doi.org/10.1109/ICASSP.1994.389932","url":null,"abstract":"We present the bootstrap sampling techniques applied to some pattern recognition algorithms. Two important procedures in image analysis are tested: a statistical segmentation based on expectation-maximisation (EM) family algorithms and two methods of invariant features extraction for gray level images. In the first case, the results we obtain show that the bootstrap sample selection method gives better results than the classical one both in the quality of the segmented image and the computing time. In the second case, the computation of the moment invariants (MI) and the analytical Fourier Mellin transform (AFMT) by the bootstrap approach using the Monte Carlo approximations are implemented. We note that this approach gives a stable approximation and reduces considerably the computing time, since we select only a small representative sample from the image. These algorithms are applied to natural image (medical image).<<ETX>>","PeriodicalId":290798,"journal":{"name":"Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115189910","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 : 1994-04-19DOI: 10.1109/ICASSP.1994.389935
D. Etter, J. Bordogna
The engineering environment of the 21st century will require new engineers to handle ambiguity and chaos, to understand the design/process/manufacture path for a product, to balance independence and teamwork, and to combine the techno-scientific base with a societal context. This paper summarizes some of the specific activities and experiments currently underway at universities in the United States to provide new educational models.<>
{"title":"Engineering education for the 21st century","authors":"D. Etter, J. Bordogna","doi":"10.1109/ICASSP.1994.389935","DOIUrl":"https://doi.org/10.1109/ICASSP.1994.389935","url":null,"abstract":"The engineering environment of the 21st century will require new engineers to handle ambiguity and chaos, to understand the design/process/manufacture path for a product, to balance independence and teamwork, and to combine the techno-scientific base with a societal context. This paper summarizes some of the specific activities and experiments currently underway at universities in the United States to provide new educational models.<<ETX>>","PeriodicalId":290798,"journal":{"name":"Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115459968","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 : 1994-04-19DOI: 10.1109/ICASSP.1994.389805
F. Nunes, J. Leitao
Microwave broadband reflectometry is used to determine the spatial distribution of plasma density (density profile) in experimental nuclear fusion devices. The position of each plasma reflecting layer is evaluated by Abel integration of the phase rate of the reflectometric signal that results from mixing the plasma incident and reflected waves. The signal processing problem is cast in the stochastic nonlinear filtering framework. The phase/frequency corresponding to the reflecting layer dynamics is modeled as a two-dimensional Markovian process, from which only one component (phase) is observed. Guided by these (noisy) observations a suboptimal filter propagates the joint (phase/frequency) a posteriori probability density function, allowing to estimate the phase rate. The purpose is to develop a robust and versatile tool to process reflectometric data.<>
{"title":"Statistical signal processing in broadband reflectometry","authors":"F. Nunes, J. Leitao","doi":"10.1109/ICASSP.1994.389805","DOIUrl":"https://doi.org/10.1109/ICASSP.1994.389805","url":null,"abstract":"Microwave broadband reflectometry is used to determine the spatial distribution of plasma density (density profile) in experimental nuclear fusion devices. The position of each plasma reflecting layer is evaluated by Abel integration of the phase rate of the reflectometric signal that results from mixing the plasma incident and reflected waves. The signal processing problem is cast in the stochastic nonlinear filtering framework. The phase/frequency corresponding to the reflecting layer dynamics is modeled as a two-dimensional Markovian process, from which only one component (phase) is observed. Guided by these (noisy) observations a suboptimal filter propagates the joint (phase/frequency) a posteriori probability density function, allowing to estimate the phase rate. The purpose is to develop a robust and versatile tool to process reflectometric data.<<ETX>>","PeriodicalId":290798,"journal":{"name":"Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122853094","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 : 1994-04-19DOI: 10.1109/ICASSP.1994.389225
E. López-Gonzalo, L. Hernández-Gómez
The aim of the proposed paper is to discuss how to model representations of both fundamental frequency and suprasegmental duration in TTS converters for Spanish. For this purpose we use a data-driven methodology that is able to represent both fundamental frequency and suprasegmental duration in order to model the prosody of a text-to-speech system for Spanish.<>
{"title":"Data-driven joint f/sub 0/ and duration modeling in text to speech conversion for Spanish","authors":"E. López-Gonzalo, L. Hernández-Gómez","doi":"10.1109/ICASSP.1994.389225","DOIUrl":"https://doi.org/10.1109/ICASSP.1994.389225","url":null,"abstract":"The aim of the proposed paper is to discuss how to model representations of both fundamental frequency and suprasegmental duration in TTS converters for Spanish. For this purpose we use a data-driven methodology that is able to represent both fundamental frequency and suprasegmental duration in order to model the prosody of a text-to-speech system for Spanish.<<ETX>>","PeriodicalId":290798,"journal":{"name":"Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing","volume":"197 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123014582","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：