Approximate Kalman filtering for the harmonic plus noise model
Authors:
- U. JainUJU. Jain
- 13 November 2002
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 75-78
Abstract
We present a probabilistic description of the harmonic plus noise model (HNM) for speech signals. This probabilistic formulation permits maximum likelihood (ML) parameter estimation and speech synthesis becomes a straightforward sampling from a distribution. It also permits the development of a Kalman filter that tracks model parameters such as pitch, harmonic amplitudes, and autoregressive coefficients. We focus here on pitch tracking for which the estimator is highly non-linear. As a result it is necessary to develop an approximate Kalman filter that goes beyond extended Kalman filtering.Keywords
This publication has 6 references indexed in Scilit:
- Applying the harmonic plus noise model in concatenative speech synthesisIEEE Transactions on Speech and Audio Processing, 2001
- Concatenative synthesis based on a harmonic modelIEEE Transactions on Speech and Audio Processing, 2001
- A Unifying Review of Linear Gaussian ModelsNeural Computation, 1999
- Maximum likelihood estimation of signals in autoregressive noiseIEEE Transactions on Signal Processing, 1994
- Extraction of periodic signals in colored noise1992
- A sound decomposition system based on a deterministic plus residual modelThe Journal of the Acoustical Society of America, 1990