The Experts below are selected from a list of 7113 Experts worldwide ranked by ideXlab platform
A. Papandreou-suppappola - One of the best experts on this subject based on the ideXlab platform.
-
Time-scale Canonical Model for wideband system characterization
Proceedings. (ICASSP '05). IEEE International Conference on Acoustics Speech and Signal Processing 2005., 2005Co-Authors: Y. Jiang, A. Papandreou-suppappolaAbstract:In this paper, we propose a time-scale Canonical Model as a discrete characterization of wideband linear time-varying systems. This representation decomposes a system output into discrete time shifts and Doppler scalings on the input, weighted by a smoothed discrete version of the wideband spreading function. We base this formulation on the Mellin transform that is matched to scalings. We also demonstrate that our proposed Model inherently affords a joint multipath-scale diversity in wideband communication channels. By properly designing the signaling and reception schemes using wavelet techniques, we can achieve this diversity over a dyadic time-scale framework.
-
ICASSP (4) - Time-scale Canonical Model for wideband system characterization
Proceedings. (ICASSP '05). IEEE International Conference on Acoustics Speech and Signal Processing 2005., 2005Co-Authors: Y. Jiang, A. Papandreou-suppappolaAbstract:In this paper, we propose a time-scale Canonical Model as a discrete characterization of wideband linear time-varying systems. This representation decomposes a system output into discrete time shifts and Doppler scalings on the input, weighted by a smoothed discrete version of the wideband spreading function. We base this formulation on the Mellin transform that is matched to scalings. We also demonstrate that our proposed Model inherently affords a joint multipath-scale diversity in wideband communication channels. By properly designing the signaling and reception schemes using wavelet techniques, we can achieve this diversity over a dyadic time-scale framework.
Y. Jiang - One of the best experts on this subject based on the ideXlab platform.
-
Time-scale Canonical Model for wideband system characterization
Proceedings. (ICASSP '05). IEEE International Conference on Acoustics Speech and Signal Processing 2005., 2005Co-Authors: Y. Jiang, A. Papandreou-suppappolaAbstract:In this paper, we propose a time-scale Canonical Model as a discrete characterization of wideband linear time-varying systems. This representation decomposes a system output into discrete time shifts and Doppler scalings on the input, weighted by a smoothed discrete version of the wideband spreading function. We base this formulation on the Mellin transform that is matched to scalings. We also demonstrate that our proposed Model inherently affords a joint multipath-scale diversity in wideband communication channels. By properly designing the signaling and reception schemes using wavelet techniques, we can achieve this diversity over a dyadic time-scale framework.
-
ICASSP (4) - Time-scale Canonical Model for wideband system characterization
Proceedings. (ICASSP '05). IEEE International Conference on Acoustics Speech and Signal Processing 2005., 2005Co-Authors: Y. Jiang, A. Papandreou-suppappolaAbstract:In this paper, we propose a time-scale Canonical Model as a discrete characterization of wideband linear time-varying systems. This representation decomposes a system output into discrete time shifts and Doppler scalings on the input, weighted by a smoothed discrete version of the wideband spreading function. We base this formulation on the Mellin transform that is matched to scalings. We also demonstrate that our proposed Model inherently affords a joint multipath-scale diversity in wideband communication channels. By properly designing the signaling and reception schemes using wavelet techniques, we can achieve this diversity over a dyadic time-scale framework.
R.w. Erickson - One of the best experts on this subject based on the ideXlab platform.
-
Extension of state-space averaging to resonant switches and beyond
IEEE Transactions on Power Electronics, 1990Co-Authors: A.f. Witulski, R.w. EricksonAbstract:It is shown that the state-space averaging method can be extended by linear network theory from the domain of pulse-width-modulated converters to a much larger class, including resonant switches and current-programmed mode. The Canonical Model concept is also extended, and it is shown that the effect of resonant switching is to introduce a feedback block into the generalized Canonical Model. These results are applied to linear zero-current and zero-voltage resonant switches, a new class of nonlinear resonant-switch converters, and the current-programmed mode. Equivalent circuit Models are developed for both full and half-wave operation, and experimental verification is presented.
D.d. Bekut - One of the best experts on this subject based on the ideXlab platform.
-
Extension of the Canonical Model to grounding parts of power systems under fault conditions
International Journal of Electrical Power & Energy Systems, 2003Co-Authors: V.c. Strezoski, Goran Svenda, D.d. BekutAbstract:This paper deals with calculations of alternate components of states of power systems in arbitrarily chosen composite fault conditions. Canonical Model is applied for these calculations. This Model was already established for calculations of states of energized parts of faulted power systems. It is applied in this paper for calculations of the states of grounded parts as well. Thus, the Model refers to arbitrarily small or large portions belonging to only one or both energized and grounded parts of power systems. By the Canonical Model application, standard procedures established for calculations of states in energized parts are extended for calculations of states of grounded parts of faulted power systems. This extension is provided by three basic generalizations: (1) the standard three-dimensional three-phase quantities are generalized as four-dimensional ones; (2) the standard three-dimensional symmetrical components transformation is generalized as four-dimensional one; (3) the application of the standard p.u. method is substituted by the generalized p.u. method. The Model is verified in details by an example referring to a line to ground short-circuit occurred on an overhead line. It is also stated for solution of a composite fault which consists of the above mentioned fault associated with a line interruption.
-
A Canonical Model for the study of faults in power systems
IEEE Transactions on Power Systems, 1991Co-Authors: V.c. Strezoski, D.d. BekutAbstract:The authors introduce a general class of fault conditions encompassing all possible faults in a power system, e.g., simultaneous symmetrical and unsymmetrical short circuits, line interruptions, and predicted breaker operations at different locations. The influence of mutually coupled lines, static VAR systems and/or faults at several busbar systems are also considered. The fault conditions are described in both the (three) phase and sequence domains, with the same simplicity. The proposed Model of a faulted power system, is in a Canonical form, over the general class of fault conditions. It is found that its effectiveness is equal in both domains. The authors also present the Model solution procedure for various fault conditions in the general frame of the study of faults. They present the derivation of the Canonical Model and numerical examples. The application of the Canonical Model to faults that include buses of distinct voltage levels is stressed.
M.j.f. Gales - One of the best experts on this subject based on the ideXlab platform.
-
Incremental Adaptation using Bayesian Inference
2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings, 2006Co-Authors: K. Yu, M.j.f. GalesAbstract:Adaptive training is a powerful technique to build system on non-homogeneous training data. Here, a Canonical Model, representing "pure" speech variability and a set of transforms representing unwanted acoustic variabilities are both trained. To use the Canonical Model for recognition, a transform for the test acoustic condition is required. For some situations a robust estimate of the transform parameters may not be possible due to limited, or no, adaptation data. One solution to this problem is to view adaptive training in a Bayesian framework and marginalise out the transform parameters. Exact implementation of this Bayesian inference is intractable. Recently, lower bound approximations based on variational Bayes have been used to solve this problem for batch adaptation with limited data. This paper extends this Bayesian adaptation framework to incremental adaptation. Various lower-bound approximations and options for propagating information within this incremental framework are discussed. Experiments using adaptive Models trained with both maximum likelihood and minimum phone error training are described. Using incremental Bayesian adaptation gains were obtained over the standard approaches, especially for limited data
-
Multiple-cluster adaptive training schemes
2001 IEEE International Conference on Acoustics Speech and Signal Processing. Proceedings (Cat. No.01CH37221), 2001Co-Authors: M.j.f. GalesAbstract:This paper examines the training of multiple-cluster systems using adaptive training schemes. Various forms of transformation and Canonical Model are described in a consistent framework allowing re-estimation formulae for all cases to be simply derived. Initial experiments using these various schemes on a large vocabulary speech recognition task are presented. The initial experiments indicate that to achieve best performance when adapting these multiple-cluster systems requires the use of adaptive training schemes rather than using simpler cluster initialisation schemes.
-
Adaptive training for robust ASR
IEEE Workshop on Automatic Speech Recognition and Understanding 2001. ASRU '01., 2001Co-Authors: M.j.f. GalesAbstract:Adaptive training is a powerful training technique for building speech recognition systems on nonhomogeneous data. The aim is to remove unwanted variability, such as changes in speaker, channel or acoustic environment, from desired changes, the acoustic differences between words. During training, two sets of Models are generated: a Canonical Model set for the desired "true" variability of the speech data, and a set of transforms to represent the unwanted variability. The Canonical Model set trained in this fashion should be more "amenable" to being adapted to a particular target condition and more "compact". During recognition, a transform to the target domain is trained. This target specific transform is then used with the Canonical Model set in the recognition process. The paper gives an overview of the underlying theory and assumptions used in adaptive training. Furthermore, the use of adaptive training schemes in current state-of-the-art tasks is described, together with a discussion of how such schemes may be used in the future.