The Experts below are selected from a list of 21534 Experts worldwide ranked by ideXlab platform
I Bazzi - One of the best experts on this subject based on the ideXlab platform.
-
tracking vocal tract resonances using a quantized nonlinear function embedded in a Temporal Constraint
IEEE Transactions on Audio Speech and Language Processing, 2006Co-Authors: Li Deng, Alex Acero, I BazziAbstract:This paper presents a new technique for high-accuracy tracking of vocal-tract resonances (which coincide with formants for nonnasalized vowels) in natural speech. The technique is based on a discretized nonlinear prediction function, which is embedded in a Temporal Constraint on the quantized input values over adjacent time frames as the prior knowledge for their Temporal behavior. The nonlinear prediction is constructed, based on its analytical form derived in detail in this paper, as a parameter-free, discrete mapping function that approximates the “forward” relationship from the resonance frequencies and bandwidths to the Linear Predictive Coding (LPC) cepstra of real speech. Discretization of the function permits the “inversion” of the function via a search operation. We further introduce the nonlinear-prediction residual, characterized by a multivariate Gaussian vector with trainable mean vectors and covariance matrices, to account for the errors due to the functional approximation. We develop and describe an expectation–maximization (EM)-based algorithm for training the parameters of the residual, and a dynamic programming-based algorithm for resonance tracking. Details of the algorithm implementation for computation speedup are provided. Experimental results are presented which demonstrate the effectiveness of our new paradigm for tracking vocal-tract resonances. In particular, we show the effectiveness of training the prediction-residual parameters in obtaining high-accuracy resonance estimates, especially during consonantal closure.
-
tracking vocal tract resonances using an analytical nonlinear predictor and a target guided Temporal Constraint
Conference of the International Speech Communication Association, 2003Co-Authors: Li Deng, I Bazzi, Alex AceroAbstract:A technique for high-accuracy tracking of formants or vocal tract resonances is presented in this paper using a novel nonlinear predictor and using a target-directed Temporal Constraint. The nonlinear predictor is constructed from a parameter-free, discrete mapping function from the formant (frequencies and bandwidths) space to the LPC-cepstral space, with trainable residuals. We examine in this study the key role of vocal tract resonance targets in the tracking accuracy. Experimental results show that due to the use of the targets, the tracked formants in the consonantal regions (including closures and short pauses) of the speech utterance exhibit the same dynamic properties as for the vocalic regions, and reflect the underlying vocal tract resonances. The results also demonstrate the effectiveness of training the prediction-residual parameters and of incorporating the target-based Constraint in obtaining high-accuracy formant estimates, especially for non-sonorant portions of speech.
Li Deng - One of the best experts on this subject based on the ideXlab platform.
-
tracking vocal tract resonances using a quantized nonlinear function embedded in a Temporal Constraint
IEEE Transactions on Audio Speech and Language Processing, 2006Co-Authors: Li Deng, Alex Acero, I BazziAbstract:This paper presents a new technique for high-accuracy tracking of vocal-tract resonances (which coincide with formants for nonnasalized vowels) in natural speech. The technique is based on a discretized nonlinear prediction function, which is embedded in a Temporal Constraint on the quantized input values over adjacent time frames as the prior knowledge for their Temporal behavior. The nonlinear prediction is constructed, based on its analytical form derived in detail in this paper, as a parameter-free, discrete mapping function that approximates the “forward” relationship from the resonance frequencies and bandwidths to the Linear Predictive Coding (LPC) cepstra of real speech. Discretization of the function permits the “inversion” of the function via a search operation. We further introduce the nonlinear-prediction residual, characterized by a multivariate Gaussian vector with trainable mean vectors and covariance matrices, to account for the errors due to the functional approximation. We develop and describe an expectation–maximization (EM)-based algorithm for training the parameters of the residual, and a dynamic programming-based algorithm for resonance tracking. Details of the algorithm implementation for computation speedup are provided. Experimental results are presented which demonstrate the effectiveness of our new paradigm for tracking vocal-tract resonances. In particular, we show the effectiveness of training the prediction-residual parameters in obtaining high-accuracy resonance estimates, especially during consonantal closure.
-
tracking vocal tract resonances using an analytical nonlinear predictor and a target guided Temporal Constraint
Conference of the International Speech Communication Association, 2003Co-Authors: Li Deng, I Bazzi, Alex AceroAbstract:A technique for high-accuracy tracking of formants or vocal tract resonances is presented in this paper using a novel nonlinear predictor and using a target-directed Temporal Constraint. The nonlinear predictor is constructed from a parameter-free, discrete mapping function from the formant (frequencies and bandwidths) space to the LPC-cepstral space, with trainable residuals. We examine in this study the key role of vocal tract resonance targets in the tracking accuracy. Experimental results show that due to the use of the targets, the tracked formants in the consonantal regions (including closures and short pauses) of the speech utterance exhibit the same dynamic properties as for the vocalic regions, and reflect the underlying vocal tract resonances. The results also demonstrate the effectiveness of training the prediction-residual parameters and of incorporating the target-based Constraint in obtaining high-accuracy formant estimates, especially for non-sonorant portions of speech.
Alex Acero - One of the best experts on this subject based on the ideXlab platform.
-
tracking vocal tract resonances using a quantized nonlinear function embedded in a Temporal Constraint
IEEE Transactions on Audio Speech and Language Processing, 2006Co-Authors: Li Deng, Alex Acero, I BazziAbstract:This paper presents a new technique for high-accuracy tracking of vocal-tract resonances (which coincide with formants for nonnasalized vowels) in natural speech. The technique is based on a discretized nonlinear prediction function, which is embedded in a Temporal Constraint on the quantized input values over adjacent time frames as the prior knowledge for their Temporal behavior. The nonlinear prediction is constructed, based on its analytical form derived in detail in this paper, as a parameter-free, discrete mapping function that approximates the “forward” relationship from the resonance frequencies and bandwidths to the Linear Predictive Coding (LPC) cepstra of real speech. Discretization of the function permits the “inversion” of the function via a search operation. We further introduce the nonlinear-prediction residual, characterized by a multivariate Gaussian vector with trainable mean vectors and covariance matrices, to account for the errors due to the functional approximation. We develop and describe an expectation–maximization (EM)-based algorithm for training the parameters of the residual, and a dynamic programming-based algorithm for resonance tracking. Details of the algorithm implementation for computation speedup are provided. Experimental results are presented which demonstrate the effectiveness of our new paradigm for tracking vocal-tract resonances. In particular, we show the effectiveness of training the prediction-residual parameters in obtaining high-accuracy resonance estimates, especially during consonantal closure.
-
tracking vocal tract resonances using an analytical nonlinear predictor and a target guided Temporal Constraint
Conference of the International Speech Communication Association, 2003Co-Authors: Li Deng, I Bazzi, Alex AceroAbstract:A technique for high-accuracy tracking of formants or vocal tract resonances is presented in this paper using a novel nonlinear predictor and using a target-directed Temporal Constraint. The nonlinear predictor is constructed from a parameter-free, discrete mapping function from the formant (frequencies and bandwidths) space to the LPC-cepstral space, with trainable residuals. We examine in this study the key role of vocal tract resonance targets in the tracking accuracy. Experimental results show that due to the use of the targets, the tracked formants in the consonantal regions (including closures and short pauses) of the speech utterance exhibit the same dynamic properties as for the vocalic regions, and reflect the underlying vocal tract resonances. The results also demonstrate the effectiveness of training the prediction-residual parameters and of incorporating the target-based Constraint in obtaining high-accuracy formant estimates, especially for non-sonorant portions of speech.
Francesca Rossi - One of the best experts on this subject based on the ideXlab platform.
-
controllability of soft Temporal Constraint problems
Lecture Notes in Computer Science, 2004Co-Authors: Francesca Rossi, Kristen Brent Venable, Neil YorkesmithAbstract:In real-life Temporal scenarios, uncertainty and preferences are often essential, coexisting aspects. We present a formalism where Temporal Constraints with both preferences and uncertainty can be defined. We show how three classical notions of controllability (strong, weak and dynamic), which have been developed for uncertain Temporal problems, can be generalised to handle also preferences. We then propose algorithms that check the presence of these properties and we prove that, in general, dealing simultaneously with preferences and uncertainty does not increase the complexity beyond that of the separate cases. In particular, we develop a dynamic execution algorithm, of polynomial complexity, that produces plans under uncertainty that are optimal w.r.t. preference.
-
Temporal Constraint reasoning with preferences
International Joint Conference on Artificial Intelligence, 2001Co-Authors: Lina Khatib, Paul Morris, Robert L Morris, Francesca RossiAbstract:A number of reasoning problems involving the manipulation of Temporal information can naturally be viewed as implicitly inducing an ordering of potential local decisions involving time (specifically, associated with durations or orderings of events) on the basis of preferences. For example. a pair of events might be constrained to occur in a certain order, and, in addition. it might be preferable that the delay between them be as large, or as small, as possible. This paper explores problems in which a set of Temporal Constraints is specified, where each Constraint is associated with preference criteria for making local decisions about the events involved in the Constraint, and a reasoner must infer a complete solution to the problem such that, to the extent possible, these local preferences are met in the best way. A Constraint framework for reasoning about time is generalized to allow for preferences over event distances and durations, and we study the complexity of solving problems in the resulting formalism. It is shown that while in general such problems are NP-hard, some restrictions on the shape of the preference functions, and on the structure of the preference set, can be enforced to achieve tractability. In these cases, a simple generalization of a single-source shortest path algorithm can be used to compute a globally preferred solution in polynomial time.
Malek Mouhoub - One of the best experts on this subject based on the ideXlab platform.
-
Specifying and solving symbolic and numeric Temporal Constraints
International Journal of Knowledge-based and Intelligent Engineering Systems, 2013Co-Authors: Samira Sadaoui, Malek MouhoubAbstract:Representing and solving combinatorial problems, especially those including Temporal Constraints, using a Constraint programming language remains a challenging task. In this paper, we present a tool to assist users in specifying and solving problems under qualitative and quantitative Temporal Constraints. The tool is based on the TemPro framework that has the ability to manage both numeric and symbolic Temporal Constraints within a unique model. Our tool provides a generic template that can be specialized to describe a wide variety of Temporal Constraint applications. Given a problem under Temporal Constraints, the proposed tool with its friendly graphical user interface first assists the user in the different steps of the problem specification. The graph representation of the Temporal Constraint problem and its consistent scenarios are then automatically generated and visualized during the solving phase. The user has also the ability to add or remove some Constraints and see the effects of these changes on the consistency of the problem.
-
Managing Temporal Constraints with Preferences
Spatial Cognition & Computation, 2008Co-Authors: Malek Mouhoub, Amrudee SukpanAbstract:Abstract Preferences in Temporal problems are common but significant in many real world applications. In this paper, we extend our Temporal reasoning framework, managing numeric and symbolic information, in order to handle preferences. Unlike the existing models managing single Temporal preferences, ours supports four types of preferences, namely: numeric and symbolic Temporal preferences, composite preferences and conditional preferences. This offers more expressive power in representing a wide variety of Temporal Constraint problems. The preferences are considered here as a set of soft Constraints using a c-semiring structure with combination and projection operators. Solving Temporal Constraint problems with preferences consists in finding a solution satisfying all the Temporal Constraints while optimizing the preference values. This is handled by a variant of the branch and bound algorithm, we propose in this paper, and where Constraint propagation is used to improve the time efficiency. Experimental ...
-
TIME - Conditional and Composite Temporal Constraints with Preferences
Thirteenth International Symposium on Temporal Representation and Reasoning (TIME'06), 2006Co-Authors: Malek Mouhoub, Amrudee SukpanAbstract:Preferences in Temporal problems are common but significant in many real world applications. In this paper, we extend our Temporal reasoning framework, managing numeric and symbolic information, in order to handle preferences. Unlike the existing models managing single Temporal preferences, ours supports four types of preferences, namely : numeric and symbolic Temporal preferences, composite preferences and conditional preferences. This offers more expressive power in representing a wide variety of Temporal Constraint problems. The preferences are considered here as a set of soft Constraints using a c-semiring structure with combination and projection operators. Solving Temporal Constraint problems with preferences consists of finding a solution satisfying all the Temporal Constraints while optimizing the preference values. This is handled by a variant of the branch and bound algorithm, we propose in this paper, and where Constraint propagation is used to improve the time efficiency. Preliminary tests, we conducted on randomly generated Temporal Constraint problems with preferences, favor the forward checking principle as a Constraint propagation strategy.
-
IASSE - Java with CREAM for Temporal Constraints.
2004Co-Authors: Malek Mouhoub, Mujtaba Istihad, Samira SadaouiAbstract:Computational problems from many different application areas can be seen as Temporal Constraint-based problems. For example, scheduling, planning, computational linguistics and database design applications can all be seen in this way. In this paper we present a Temporal Constraint solver based on the Java Cream Constraint library for managing problems involving numeric and symbolic Temporal information. The solving system comes with a graphic user interface that allows the user to input the Temporal information of a given problem and to check the consistency of these Constraints in an interactive manner.
-
Solving Temporal Constraints in Real Time and in a Dynamic Environment
2002Co-Authors: Malek MouhoubAbstract:In this paper we will present a study of different resolution techniques for solving Constraint Satisfaction Problems(CSP) in the case of Temporal Constraints. This later problem is called Temporal Constraint Satisfaction Problem(TCSP). We will mainly focus here on solving TCSPs in real time and in a dynamic environment. Indeed, addressing these two issues is very relevant for many real world applications. Solving a TCSP in real time is an optimization problem that we call MTCSP(Maximal Temporal Constraint Satisfaction Problems). The objective function to minimize is the number of Temporal Constraint violations. The results of the tests we have performed on randomly generated MTCSPs show that the approximation method Min-Conflict-Random-Walk(MCRW) is the algorithm of choice for solving MTCSPs. Comparison study of the different dynamic arc-consistency algorithms for solving dynamicTemporal Constraint problems in a preprocessing phase demonstrates that the new algorithm we propose and based on a recent arc-consistency algorithm represents a better compromise between time and space than the other dynamic arc-consistency algorithms.
