Spectral envelope estimation using a penalized likelihood criterion

Olivier Cappe; Marine Oudot; Eric Moulines

Spectral envelope estimation using a penalized likelihood criterion

Olivier Cappe
, Marine Oudot
, Eric Moulines

Centre national de la recherche scientifique

Research output: Contribution to conference › Paper › peer-review

Abstract

A penalized likelihood criterion is introduced for estimating the spectral envelope in the presence of measurement noise. A performance criterion is developed for assessing the goodness of fit of the envelope which is motivated by the statistical analysis of the model under consideration. This criterion is obtained as a large sample approximation of the likelihood for the estimated values |H(ω-k$/)|². The envelope itself is estimated by numerical optimization of a penalized version of the criterion.

Original language	English
Publication status	Published - 1 Dec 1997
Event	Proceedings of the 1997 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA - New Paltz, NY, USA Duration: 19 Oct 1997 → 22 Oct 1997

Conference

Conference	Proceedings of the 1997 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA
City	New Paltz, NY, USA
Period	19/10/97 → 22/10/97

Cite this

@conference{19ce219af90f48779bf857e3b3b08881,

title = "Spectral envelope estimation using a penalized likelihood criterion",

abstract = "A penalized likelihood criterion is introduced for estimating the spectral envelope in the presence of measurement noise. A performance criterion is developed for assessing the goodness of fit of the envelope which is motivated by the statistical analysis of the model under consideration. This criterion is obtained as a large sample approximation of the likelihood for the estimated values |H(ω-k\$/)|2. The envelope itself is estimated by numerical optimization of a penalized version of the criterion.",

author = "Olivier Cappe and Marine Oudot and Eric Moulines",

year = "1997",

month = dec,

day = "1",

language = "English",

note = "Proceedings of the 1997 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA ; Conference date: 19-10-1997 Through 22-10-1997",

}

TY - CONF

T1 - Spectral envelope estimation using a penalized likelihood criterion

AU - Cappe, Olivier

AU - Oudot, Marine

AU - Moulines, Eric

PY - 1997/12/1

Y1 - 1997/12/1

N2 - A penalized likelihood criterion is introduced for estimating the spectral envelope in the presence of measurement noise. A performance criterion is developed for assessing the goodness of fit of the envelope which is motivated by the statistical analysis of the model under consideration. This criterion is obtained as a large sample approximation of the likelihood for the estimated values |H(ω-k$/)|2. The envelope itself is estimated by numerical optimization of a penalized version of the criterion.

AB - A penalized likelihood criterion is introduced for estimating the spectral envelope in the presence of measurement noise. A performance criterion is developed for assessing the goodness of fit of the envelope which is motivated by the statistical analysis of the model under consideration. This criterion is obtained as a large sample approximation of the likelihood for the estimated values |H(ω-k$/)|2. The envelope itself is estimated by numerical optimization of a penalized version of the criterion.

UR - https://www.scopus.com/pages/publications/0031379141

M3 - Paper

AN - SCOPUS:0031379141

T2 - Proceedings of the 1997 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA

Y2 - 19 October 1997 through 22 October 1997

ER -

Spectral envelope estimation using a penalized likelihood criterion

Abstract

Conference

Other files and links

Fingerprint

Cite this