Catalogo Articoli (Spogli Riviste)

OPAC HELP

Titolo:
Power exponential densities for the training and classification of acoustic feature vectors in speech recognition
Autore:
Basu, S; Micchelli, CA; Olsen, P;
Indirizzi:
IBM Corp, Thomas J Watson Res Ctr, Yorktown Heights, NY 10598 USA IBM Corp Yorktown Heights NY USA 10598 tr, Yorktown Heights, NY 10598 USA SUNY Albany, Dept Math & Stat, Albany, NY 12222 USA SUNY Albany Albany NYUSA 12222 y, Dept Math & Stat, Albany, NY 12222 USA
Titolo Testata:
JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS
fascicolo: 1, volume: 10, anno: 2001,
pagine: 158 - 184
SICI:
1061-8600(200103)10:1<158:PEDFTT>2.0.ZU;2-1
Fonte:
ISI
Lingua:
ENG
Keywords:
digamma function; expectation maximization algorithm; Laplacian densities; large vocabulary continuous speech recognition; univariate and multivariate density estimation;
Tipo documento:
Article
Natura:
Periodico
Settore Disciplinare:
Physical, Chemical & Earth Sciences
Citazioni:
20
Recensione:
Indirizzi per estratti:
Indirizzo: Basu, S IBM Corp, Thomas J Watson Res Ctr, Yorktown Heights, NY 10598 USA IBM Corp Yorktown Heights NY USA 10598 town Heights, NY 10598 USA
Citazione:
S. Basu et al., "Power exponential densities for the training and classification of acoustic feature vectors in speech recognition", J COMPU G S, 10(1), 2001, pp. 158-184

Abstract

We consider a parametric family of multivariate density functions formed by mixture models from univariate functions of the type exp(-\x\(alpha)) formodeling acoustic feature vectors used in automatic recognition of speech. The parameter alpha is used to measure the non-Gaussian nature of the data. Previous work has focused on estimating the mean and the variance of the data for a fixed alpha. Here we attempt to estimate the alpha from the datausing a maximum likelihood criterion. Among other things, we show that there is a balance between a and the number of data points N that must be satisfied for efficient estimation. Numerical experiments are performed on multidimensional vectors obtained from speech data.

ASDD Area Sistemi Dipartimentali e Documentali, Università di Bologna, Catalogo delle riviste ed altri periodici
Documento generato il 05/04/20 alle ore 03:12:14