Time-Varying Autoregressions in Speech: Detection Theory and Applications

Rudoy, Daniel; Quatieri, Thomas F.; Wolfe, Patrick J.

doi:10.1109/tasl.2010.2073704

Cited by 30 publications

(25 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, it is noted that all the linear predictive analyses involved in the present study are classical in the sense that they use time-invariant filter coefficients that are updated once per frame. A more flexible paradigm is to utilize time-varying AR-modeling (e.g., Schnell and Lacroix, 2008;Rudoy et al, 2011) in which linear predictive filter coefficients evolve in time. Combining the proposed WLP-AME method with the time-varying AR modeling approach is another topic of future studies which would maybe help in detecting vocal tract variation in continuous high-pitched speech.…”

Section: Discussionmentioning

confidence: 99%

Formant frequency estimation of high-pitched vowels using weighted linear prediction

Alku

Pohjalainen

Vainio

et al. 2013

The Journal of the Acoustical Society of America

View full text Add to dashboard Cite

All-pole modeling is a widely used formant estimation method, but its performance is known to deteriorate for high-pitched voices. In order to address this problem, several all-pole modeling methods robust to fundamental frequency have been proposed. This study compares five such previously known methods and introduces a technique, Weighted Linear Prediction with Attenuated Main Excitation (WLP-AME). WLP-AME utilizes temporally weighted linear prediction (LP) in which the square of the prediction error is multiplied by a given parametric weighting function. The weighting downgrades the contribution of the main excitation of the vocal tract in optimizing the filter coefficients. Consequently, the resulting all-pole model is affected more by the characteristics of the vocal tract leading to less biased formant estimates. By using synthetic vowels created with a physical modeling approach, the results showed that WLP-AME yields improved formant frequencies for high-pitched sounds in comparison to the previously known methods (e.g., relative error in the first formant of the vowel [a] decreased from 11% to 3% when conventional LP was replaced with WLP-AME). Experiments conducted on natural vowels indicate that the formants detected by WLP-AME changed in a more regular manner between repetitions of different pitch than those computed by conventional LP.

show abstract

Section: Discussionmentioning

confidence: 99%

Formant frequency estimation of high-pitched vowels using weighted linear prediction

Alku

Pohjalainen

Vainio

et al. 2013

The Journal of the Acoustical Society of America

View full text Add to dashboard Cite

show abstract

“…In this study, we propose to modify the 2DAR model by replacing TDLP with time-varying linear prediction (TVLP) [11,12] which is a generalization of conventional linear prediction (LP) [13]. TVLP can be used to analyze non-stationarity of speech signals by allowing the underlying all-pole model to be time-varying.…”

Section: Introductionmentioning

confidence: 99%

Time-Varying Autoregressions for Speaker Verification in Reverberant Conditions

Vestman¹,

Gowda²,

Sahidullah³

et al. 2017

Interspeech 2017

View full text Add to dashboard Cite

In poor room acoustics conditions, speech signals received by a microphone might become corrupted by the signals' delayed versions that are reflected from the room surfaces (e.g. wall, floor). This phenomenon, reverberation, drops the accuracy of automatic speaker verification systems by causing mismatch between the training and testing. Since reverberation causes temporal smearing to the signal, one way to tackle its effects is to study robust feature extraction, particularly based on long-time temporal feature extraction. This approach has been adopted previously in the form of 2-dimensional autoregressive (2DAR) feature extraction scheme by using frequency domain linear prediction (FDLP). In 2DAR, FDLP processing is followed by time domain linear prediction (TDLP). In the current study, we propose modifying the latter part of the 2DAR feature extraction scheme by replacing TDLP with time-varying linear prediction (TVLP) to add an extra layer of temporal processing. Our speaker verification experiments using the proposed features with the text-dependent RedDots corpus show small but consistent improvements in clean and reverberant conditions (up to 6.5%) over the 2DAR features and large improvements over the MFCC features in reverberant conditions (up to 46.5%).

show abstract

“…The other automatic segmentation group consists of algorithms which do not require any knowledge about the phonetic content and are based mostly on statistical signal analysis (e.g., Tyagi et al 2006;Almanidis et al, 2008Almanidis et al, , 2009Scharenborg et al, 2010;Rudoy et al, 2011). Tyagi et al (2006) employed speech signal modelling based on the autoregressive process.…”

Section: Introductionmentioning

confidence: 99%

“…In the work by Scharenborg et al (2010), the first step is also MFCC parameterisation, followed by segmentation based on the principle of finding the maximum distances between observation vectors in the selected subset, performed with the use of a method known as Maximum Margin Clustering (MMC). Finally, Rudoy et al (2011) used stochastic modelling employing the standard AutoRegressive (AR) and Time-Varying AutoRegressive (TVAR) models. Detection is performed with the use of a classic GLRT test based on determining which model is more adequate for a particular segment of the signal.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Automatic speech signal segmentation based on the innovation adaptive filter

Makowski

Hossa

2014

International Journal of Applied Mathematics and Computer Science

View full text Add to dashboard Cite

Speech segmentation is an essential stage in designing automatic speech recognition systems and one can find several algorithms proposed in the literature. It is a difficult problem, as speech is immensely variable. The aim of the authors' studies was to design an algorithm that could be employed at the stage of automatic speech recognition. This would make it possible to avoid some problems related to speech signal parametrization. Posing the problem in such a way requires the algorithm to be capable of working in real time. The only such algorithm was proposed by Tyagi et al., (2006), and it is a modified version of Brandt's algorithm. The article presents a new algorithm for unsupervised automatic speech signal segmentation. It performs segmentation without access to information about the phonetic content of the utterances, relying exclusively on second-order statistics of a speech signal. The starting point for the proposed method is time-varying Schur coefficients of an innovation adaptive filter. The Schur algorithm is known to be fast, precise, stable and capable of rapidly tracking changes in second order signal statistics. A transfer from one phoneme to another in the speech signal always indicates a change in signal statistics caused by vocal track changes. In order to allow for the properties of human hearing, detection of inter-phoneme boundaries is performed based on statistics defined on the mel spectrum determined from the reflection coefficients. The paper presents the structure of the algorithm, defines its properties, lists parameter values, describes detection efficiency results, and compares them with those for another algorithm. The obtained segmentation results, are satisfactory.

show abstract

Time-Varying Autoregressions in Speech: Detection Theory and Applications

Cited by 30 publications

References 47 publications

Formant frequency estimation of high-pitched vowels using weighted linear prediction

Formant frequency estimation of high-pitched vowels using weighted linear prediction

Time-Varying Autoregressions for Speaker Verification in Reverberant Conditions

Automatic speech signal segmentation based on the innovation adaptive filter

Contact Info

Product

Resources

About