On Optimal Filtering for Speech Decomposition

Jaramillo, Alfredo Esquivel; Nielsen, Jesper Kjær; Christensen, Mads Græsbøll

doi:10.23919/eusipco.2018.8553512

Cited by 2 publications

(3 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The problem of decomposing speech into its voiced and unvoiced components is useful in applications such as speech coding, analysis, synthesis, modification and diagnosing of illnesses [1][2][3][4][5][6][7]. As implied by hybrid speech models (e.g., harmonic plus noise model) [6,8], deterministic and stochastic components may coexist in a speech segment.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Speech Decomposition Based on a Hybrid Speech Model and Optimal Segmentation

Jaramillo¹,

Nielsen²,

Christensen

2021

Interspeech 2021

Self Cite

View full text Add to dashboard Cite

In a hybrid speech model, both voiced and unvoiced components can coexist in a segment. Often, the voiced speech is regarded as the deterministic component, and the unvoiced speech and additive noise are the stochastic components. Typically, the speech signal is considered stationary within fixed segments of 20-40 ms, but the degree of stationarity varies over time. For decomposing noisy speech into its voiced and unvoiced components, a fixed segmentation may be too crude, and we here propose to adapt the segment length according to the signal local characteristics. The segmentation relies on parameter estimates of a hybrid speech model and the maximum a posteriori (MAP) and log-likelihood criteria as rules for model selection among the possible segment lengths, for voiced and unvoiced speech, respectively. Given the optimal segmentation markers and the estimated statistics, both components are estimated using linear filtering. A codebook-based approach differentiates between unvoiced speech and noise. A better extraction of the components is possible by taking into account the adaptive segmentation, compared to a fixed one. Also, a lower distortion for voiced speech and higher segSNR for both components is possible, as compared to other decomposition methods.

show abstract

Section: Introductionmentioning

confidence: 99%

“…An effort based on linear filtering to estimate separately the voiced and unvoiced parts from noisy speech was presented in [3]. In this paper, instead of relying on conventional noise tracking methods (e.g., [19]), the noise statistics were estimated using approaches which rely on prior spectral information contained in codebooks [20,21].…”

Section: Introductionmentioning

confidence: 99%

Speech Decomposition Based on a Hybrid Speech Model and Optimal Segmentation

Jaramillo¹,

Nielsen²,

Christensen

2021

Interspeech 2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…The problem of estimating the fundamental frequency (a.k.a. pitch) of a periodic signal has received considerable attention during recent decades, and is of particular importance in many forms of audio and speech processing, such as speaker identification [1], audio coding [2], music transcription [3], and speech decomposition [4]. As opposed to correlation-based methods (e.g.…”

Section: Introductionmentioning

confidence: 99%

Robust Fundamental Frequency Estimation in Coloured Noise

Jaramillo

Jakobsson

Nielsen

et al. 2020

ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

Most parametric fundamental frequency estimators make the implicit assumption that any corrupting noise is additive, white Gaussian. Under this assumption, the maximum likelihood (ML) and the least squares estimators are the same, and statistically efficient. However, in the coloured noise case, the estimators differ, and the spectral shape of the corrupting noise should be taken into account. To allow for this, we here propose two schemes that refine the noise statistics and parameter estimates in an iterative manner, one of them based on an approximate ML solution and the other one based on removing the periodic signal obtained from a linearly constrained minimum variance (LCMV) filter. Evaluations on real speech data indicate that the iteration steps improve the estimation accuracy, therefore offering improvement over traditional non-parametric fundamental frequency methods in most of the evaluated scenarios.

show abstract

On Optimal Filtering for Speech Decomposition

Cited by 2 publications

References 27 publications

Speech Decomposition Based on a Hybrid Speech Model and Optimal Segmentation

Speech Decomposition Based on a Hybrid Speech Model and Optimal Segmentation

Robust Fundamental Frequency Estimation in Coloured Noise

Contact Info

Product

Resources

About