|Title:||Covariance matrix enhancement approach to train robust Gaussian mixture models of speech data|
Psutka, Josef V.
|Citation:||VANÄšK, Jan; MACHLICA, LukĂˇĹˇ; PSUTKA, Josef V.; PSUTKA, Josef. Covariance matrix enhancement approach to train robust Gaussian mixture models of speech data. In: Speech and computer. Berlin: Springer, 2013, p. 92-99. (Lectures notes in computer science; 8113). ISBN 978-3-319-01930-7.|
|Keywords:||smÄ›si GaussovskĂ˝ch modelĹŻ;plnĂˇ kovariance;plnĂ© kovarianÄŤnĂ matice;regularizace;automatickĂ© rozpoznĂˇvĂˇnĂ Ĺ™eÄŤi|
|Keywords in different language:||Gaussian mixture models;full covariance;full covariance matrix;regularization;automatic speech recognition|
|Abstract in different language:||An estimation of parameters of a multivariate Gaussian Mixture Model is usually based on a criterion (e.g. Maximum Likelihood) that is focused mostly on training data. Therefore, testing data, which were not seen during the training procedure, may cause problems. Moreover, numerical instabilities can occur (e.g. for low-occupied Gaussians especially when working with full-covariance matrices in high-dimensional spaces). Another question concerns the number of Gaussians to be trained for a specific data set. The approach proposed in this paper can handle all these issues. It is based on an assumption that the training and testing data were generated from the same source distribution. The key part of the approach is to use a criterion based on the source distribution rather than using the training data itself. It is shown how to modify an estimation procedure in order to fit the source distribution better (despite the fact that it is unknown), and subsequently new estimation algorithm for diagonal- as well as full-covariance matrices is derived and tested.|
|Rights:||Â© Jan VanÄ›k - LukĂˇĹˇ Machlica - Josef V. Psutka - Josef Psutka|
|Appears in Collections:||Články / Articles (KIV)|
Články / Articles (KKY)
Files in This Item:
|JanVanek_2013_CovarianceMatrix.pdf||Plný text||253,94 kB||Adobe PDF||View/Open|
Please use this identifier to cite or link to this item:
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.