Composite autoregressive system for sparse source-filter representation of speech

This paper presents a new generative model for speech signals called a ldquocomposite autoregressive systemrdquo. This model consists of a composite dictionary incorporating a set of the power spectral densities (PSDs) of excitation sources and a set of all-pole filters where the gain of each pair o...

Full description

Saved in:
Bibliographic Details
Published in2009 IEEE International Symposium on Circuits and Systems (ISCAS) pp. 2477 - 2480
Main Authors Kameoka, H., Kashino, K.
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.05.2009
Subjects
Dictionaries
Iterative algorithms
Nonlinear filters
Power system modeling
Signal analysis
Signal generators
Sparse matrices
Speech analysis
Speech processing
Vectors
Online AccessGet full text

Cover

More Information
Summary:This paper presents a new generative model for speech signals called a ldquocomposite autoregressive systemrdquo. This model consists of a composite dictionary incorporating a set of the power spectral densities (PSDs) of excitation sources and a set of all-pole filters where the gain of each pair of excitation and filter elements is allowed to vary over time. We use this model to develop a computationally efficient scheme for generating a sparse mixture representation of speech based on the Expectation-Maximization algorithm. The algorithm iteratively updates the excitation PSDs and the gains through the update formulae, which reduce under a particular condition to the multiplicative update rule for non-negative matrix factorization with the Itakura-Saito distance criterion, and the all-pole parameters using the Levinson-Durbin algorithm.
ISBN:1424438276
9781424438273
ISSN:0271-4302
2158-1525
DOI:10.1109/ISCAS.2009.5118303