Conference paper
Multi-Pitch Estimation using Harmonic MUSIC
In this paper, we present a method for estimation of the fundamental frequencies, or pitches, of several periodic sources. This difficult estimation problem occurs, for example, in speech and audio processing whenever multiple speakers or tones are present. The presented method is an extension of a recently proposed method for estimation of the fundamental frequency of a set of harmonically related sinusoids based on the MUSIC criterion.
The estimator takes into account that the number of harmonics and sources may vary. The performance of the proposed method is evaluated in Monte Carlo simulations using synthetic signals under various conditions. The work of M. G. Christensen is supported by the Intelligent Sound project, Danish Technical Research Council grant no. 26-04-0092.
Language: | English |
---|---|
Year: | 2006 |
Pages: | 521-524 |
ISBN: | 1424407842 , 1424407850 , 9781424407842 and 9781424407859 |
ISSN: | 10586393 and 25762303 |
Types: | Conference paper |
DOI: | 10.1109/ACSSC.2006.354802 |
Additive noise Amplitude estimation Cost function Covariance matrix Data models Frequency estimation Gaussian noise Multiple signal classification Parameter estimation Speech processing frequency estimation harmonic multiple signal classification multi pitch estimation periodic source signal classification