Energy separation in signal modulations with application to speech analysis

Maragos, Petros; Kaiser, J.; Quatieri, Thomas F.

doi:10.1109/78.277799

Cited by 683 publications

(463 citation statements)

References 31 publications

Supporting

Mentioning

430

Contrasting

Unclassified

Order By: Relevance

“…Numerous methods have been proposed [16][17][18][19] that attempt to exploit in some way turbulence related phenomena of the speech production system. Work in this area includes the application of fractal measures on the analysis of speech signals [20,2], application of nonlinear oscillator models to speech modeling, prediction and synthesis [21][22][23], or multifractal aspects [24].…”

Section: Introductionmentioning

confidence: 99%

Analysis and classification of speech signals by generalized fractal dimension features

Pitsikalis

Maragos

2009

Speech Communication

View full text Add to dashboard Cite

We explore nonlinear signal processing methods inspired by dynamical systems and fractal theory in order to analyze and characterize speech sounds. A speech signal is at first embedded in a multidimensional phase-space and further employed for the estimation of measurements related to the fractal dimensions. Our goals are to compute these raw measurements in the practical cases of speech signals, to further utilize them for the extraction of simple descriptive features and to address issues on the efficacy of the proposed features to characterize speech sounds. We observe that distinct feature vector elements obtain values or show statistical trends that on average depend on general characteristics such as the voicing, the manner and the place of articulation of broad phoneme classes. Moreover the way that the statistical parameters of the features are altered as an effect of the variation of phonetic characteristics seem to follow some roughly formed patterns. We also discuss some qualitative aspects concerning the linear phoneme-wise correlation between the fractal features and the commonly employed mel-frequency cepstral coefficients (MFCC) demonstrating phonetic cases of maximal and minimal correlation. In the same context we also investigate the fractal features' spectral content, in terms of the most and least correlated components with the MFCC.

show abstract

Section: Introductionmentioning

confidence: 99%

Analysis and classification of speech signals by generalized fractal dimension features

Pitsikalis

Maragos

2009

Speech Communication

View full text Add to dashboard Cite

show abstract

“…Teager was convinced that the speech resonances can change rapidly both in frequency and amplitude even within a single pitch period, possibly due to separated air ow in the vocal tract [5]. The nonlinear di erential Teager energy operator can detect formant AM-FM modulations by estimating the product of their time-varying amplitude and frequency.…”

Section: Teager Energy Operator Principlesmentioning

confidence: 99%

“…The Teager operator is considered a high-resolution energy estimator. The AM-FM model proposed by Maragos, Kaiser and Quatieri [5] represents a single speech resonance R(t) as an AM-FM signal…”

Section: Teager Energy Operator Principlesmentioning

confidence: 99%

“…The instantaneous formant frequency of the signal is de ned as fi(t) = f c + q ( t ). To demodulate a speech resonance R(t) i n to its varying amplitude ja(t)j and instantaneous frequency fi(t), the energy separation algorithm (ESA) (developed by Maragos, Kaiser and Quatieri [5]) is applied to the signal resonance R(t) obtained after ltering the speech signal around the formant under consideration. The ESA is based on the Teager-Kaiser energy tracking operator, with discrete time domain representation of the form,…”

Section: Teager Energy Operator Principlesmentioning

confidence: 99%

“…and the ESA frequency and amplitude estimates using the DESA-2 algorithm (Maragos, Kaiser and Quatieri [5]) are…”

Section: Teager Energy Operator Principlesmentioning

confidence: 99%

See 2 more Smart Citations

Vocal fold pathology assessment using AM autocorrelation analysis of the Teager energy operator

Gavidia-Ceballos

Hansen

Kaiser

Proceeding of Fourth International Conference on Spoken Language Processing. ICSLP '96

Self Cite

View full text Add to dashboard Cite

Traditional speech processing methods for laryngeal pathology assessment assume linear speech production, with measures derived f r om an estimated glottal ow waveform. They normally require the speaker to achieve complete glottal closure, which for many vocal fold pathologies c annot be a c c omplished. To address this, a nonlinear signal processing approach is proposed which employs a di erential Teager energy operator and the energy separation algorithm to obtain formant AM and FM modulations from bandpass ltered s p e e ch recordings. A new speech measure i s p r oposed b ased o n p arameterization of the autocorrelation envelop of the AM response. Using a cubic model of the autocorrelation envelop, a three dimensional space is formed to assess changes in speech quality. This approach is shown to achieve exemplary detection performance for a set of muscular tension dysphonias. Unlike ow characterization using numerical solutions of Navier Stokes equations, this method i s e xtremely computationally attractive, requiring only N logN +8 N multiplications and N square r o ots for N samples, and is therefore suitable for real time applications due to its computational simplicity. The new non-invasive method shows conclusively that a fast, e ective digital speech processing technique can be developed for vocal fold pathology assessment, without the need for (i) direct glottal ow estimation or (ii) complete glottal closure by the speaker. The proposed method also con rms that alternative nonlinear methods can begin to address the limitations of previous linear approaches for speech pathology assessment.

show abstract

Online tracking of voltage flicker for inverter‐based distributed generation using Teager energy operator

Karimian

Hosseinian

2019

Int Trans Electr Energ Syst

View full text Add to dashboard Cite

Summary Nowadays, increasing nonlinear loads as well as distributed generation (DG) sources, which are some of these types, not only cause a wide range of voltage fluctuations, but also make it difficult to control the DG sources to produce proper output. Hence, it is necessary to use an approach that evaluates and tracks the voltage fluctuations, and also to have an output with optimal accuracy and smoothness that is suitable for control of DG sources. This paper presents a Teager energy operator (TEO)‐based approach to track and detect the voltage fluctuations in distribution networks in the presence of the inverter‐based DG and nonlinear load that eliminates as much as possible distortion and unwanted signals from the voltage flicker signal. This hybrid algorithm is able to track accurate the amplitude variations of voltage disturbances in industrial systems. To attain this aim, a suitable combination of the TEO and a relevant filter set is adopted. Simulations for a mathematical fluctuation waveform and the grid‐connected inverter‐based DG using the proposed method are conducted and results are compared with each other to evaluate their performance. Simulation results for different voltage fluctuations validate that the new approach is accurate and smooth in tracking the voltage fluctuations.

show abstract

Energy separation in signal modulations with application to speech analysis

Cited by 683 publications

References 31 publications

Analysis and classification of speech signals by generalized fractal dimension features

Analysis and classification of speech signals by generalized fractal dimension features

Vocal fold pathology assessment using AM autocorrelation analysis of the Teager energy operator

Online tracking of voltage flicker for inverter‐based distributed generation using Teager energy operator

Contact Info

Product

Resources

About