Dale E. Veeneman

Automatic glottal inverse filtering from speech and electroglottographic signals

The flow of air through the glottis, the glottal volume-velocity, reflects the action of the vocal folds and is thus an important indicator of laryngeal function. However, it cannot be measured directly because of vocal tract filtering. We have developed an automated on-line method to determine the glottal volume-velocity waveform from normal and pathological speech based on digital inverse filtering. The method developed addresses the problems of accurate identification of vocal tract parameters and reduction of low-frequency noise. The vocal tract filter is estimated by analysis of the undriven vocal tract response during closed glottis, as identified from an electroglottographic signal. Low-frequency noise is attenuated by a high-pass filtering operation followed by a low-pass compensation. The complete inverse filtering method provides reliable glottal volume-velocity waveforms for both normal and pathological speech.

Efficient Multi-Tap Pitch Prediction for Stochastic Coding

In addition to the codebook excitation, the pitch or long-term predictor is of critical importance in determining the quality of the reconstructed speech in stochastic or CELP coding. Of equal concern is that when the pitch predictor is used in a “closed-loop” configuration, it has a high computational complexity and consumes, with the codebook search, a major portion of the coder’s computational requirement. While a higher-order predictor (e.g., 3 adjacent taps) provides improved performance (primarily because of the implicit interpolated non-integer value for the effective lag), it also requires an increase in complexity (especially when performing a closed-loop analysis) and an increase in bit-rate. However, it is possible to reduce the complexity of a 3-tap filter to be nearly comparable to a 1-tap filter and use a moderate increase in bit-rate to gain an increase in performance.

A fully adaptive comb filter for enhancing block-coded speech

Framing noise is a distortion present in block-coded speech that is caused by discontinuities at frame boundaries. A comb filter that is adapted nonpitch-synchronously has been developed that it is very effective in reducing this distortion. An optimal, minimum mean-squared-error approach is applied to determine the filter coefficients and adapt them to the short-time statistics of the speech. This filter has also proven effective in enhancing speech corrupted by other noise sources. >

An efficient code structure and search strategy for stochastic coding at 8 kb/s

Methods of increasing the performance of two components of stochastic coding, the codebook excitation search and the long-term synthesis filter, while maintaining or reducing complexity are discussed. A flexible tree-coding technique is introduced that significantly lowers the complexity of the codebook search with very little loss in quality. An efficient method of performing a three-tap long-term prediction analysis in a closed loop is developed. It improves perceptual quality and requires little additional complexity. The result is a stochastic coder which produces high-quality speech at 8 kb/s with manageable computational complexity.<<ETX>>

Automatic glottal inverse filtering

Early detection or abnormal laryngeal function is generally difficult with present techniques. The glottal volume-velocity (v-v) is an indicator of laryngeal function, but cannot be measured directly from acoustic speech because of vocal tract filtering. Digital inverse filtering is the basis for an automatic on-line method we have developed to determine the glottal v-v for normal and pathological speech. This method was developed by addressing four specific areas. First, the interval of the undriven vocal tract response during closed glottis is identified from an electroglottographic signal. Second, the vocal tract configuration is estimated by a linear prediction analyis of the closed glottis interval of the speech waveform. Third, low frequency noise is removed by a high pass filtering operation followed by a low pass compensation. Fourth, the low frequency, bias inherent in the process is removed.

Computationally efficient stochastic coding of speech

Issues of codebook excitation and the long-term synthesis filter are addressed. To deal with the excitation complexity, a codebook structure and an alternative search technique are developed for the excitation codebook that significantly lower the complexity with very little loss in quality and compare favorably with other recent alternative techniques. This allows a reasonable implementation complexity with very little loss in quality compared to a standard exhaustively-searched block codebook. An efficient method to improve the long-term prediction analysis by using a three-tap filter that requires little additional complexity is introduced. The result is a stochastic coder that provides high speech quality at 8 kb/s with manageable computational complexity.<<ETX>>

Comments on "Pole-Zero Modeling and Classification of Phonocardiograms"

Recently,1 the Steiglitz-McBride method was used to identify the two most dominant frequencies in aortic heart sounds from porcine prosthetic heart valves. A quadratic classifier was developed to evaluate valve sounds in a test set. We have corrected computational errors, recomputed their results, and interchanged the training and test sets to yield new results.