At&t Bell Laboratories

American Telephone & Telegraph Co., At&t Bell Laboratories

American Telephone and Telegraph, Co.

Dimitrios P Prezas

Walter T Hartwell

Joseph Picone

Willem Bastiaan Kleijn

David Lynn Thomson

Gustavus H Zimmerman, Iii

Edward C Bronson

Nancy M Saraf

An speech analysis and synthesis system where pitch information for excitation is transmitted during voiced segments of speech and modified residual information for excitation is transmitted during unvoiced speech segments along with linear predictive coded (LPC) parameters. The speech analysis portion of the system uses a pitch detection circuit to determine when the speech is voiced or unvoiced and to calculate the pitch information during voiced segments. A multi-pulse excitation forming circuit generates the modified residual signal which is obtained from the cross correlation of the residual signal and the LPC-recreated original signal. The pitch detection circuit controls a multiplexer which selects either the output of the multi-pulse excitation forming circuit or the output of the pitch detection circuit for transmission as the excitation information with LPC parameters to the synthesizer portion of the system.

Digital speech coder with different excitation types

A speech analysis and synthesis system where pitch information for excitation is transmitted during voice segments of speech and pulse excitation or noise excitation is transmitted during unvoiced speech segments along with linear predictive coding (LPC) parameters. The decision of whether to transmit noise excitation or pulse excitation is performed by comparing the variance of the residual to the square of the mean amplitude of the rectified residual for each frame. If the result of this comparison is greater than a threshold value, pulse excitation is utilized otherwise noise excitation is used. The pulse excitation comprises a subset of samples of the LPC residual as determined by the relative amplitudes and spacing of the local maxima in the LPC residual.

Voice synthesis utilizing multi-level filter excitation

A pitch detector system for use with speech analysis and synthesis methods having a plurality of identical detectors each responsive to a different portion of a speech signal for estimating a pitch value and a voter circuit responsive to the estimated pitch values for determining a final pitch value. The pitch detectors are identical in design which allows for an efficient software implementation since only one set of program instructions is necessary to implement all of the encoders. The voter subsystem may be implemented by a digital signal processor executing program instructions that calculate a pitch value from the estimated pitch values determined by the pitch detectors and a second set of program instructions for constraining the final pitch value outputted by the voter subsystem so that the calculated pitch value is in agreement with calculated pitch values for previous frames. In addition, the pitch and voters may be implemented by a third set of program instructions executing on the same digital signal processor as the sets of instructions for the voter subsystem.

Parallel processing pitch detector

A speech analyzer and synthesizer system using sinusoidal encoding and decoding techniques for voiced frames and noise excitation or multiple pulse excitation for unvoiced frames. For voiced frames, the analyzer transmits the pitch, values for each harmonic frequency by defining the offset from integer multiples of the fundamental frequency, total frame energy, and linear predictive coding, LPC, coefficients. The synthesizer is responsive to that information to determine the phase of the fundamental frequency and each harmonic based on the transmitted pitch and harmonic offset information and to determine the amplitudes of the harmonics utilizing the total frame energy and LPC coefficients. Once the phase and amplitudes have been determined for the fundamental and harmonic frequencies, the sinusoidal analysis is performed for voiced frames. For each frame, the determined frequencies and amplitudes are defined at the center of the frame, and a linear interpolation is used both to determine continuous frequency and amplitude signals of the fundamental and the harmonics throughout the entire frame by the synthesizer. In addition, the analyzer initially adjusts the pitch so that the harmonics are evenly distributed around integer multiples of this pitch.

Digital speech vocoder

A system for determining the heart rate in response to an electrocardiogram (EKG) signal that has been filtered by performing an autoregressive analysis, low-pass filtered by performing a digital filtering function, and the rate determined by utilizing a digital peak detection technique. In a heart rate detection system, the EKG signal from an EKG instrument is first digitized and then the colored noise component is removed from the digitized EKG samples by performing a linear predictive coding filtering function on the EKG samples. After the colored noise has been filtered from the digitized EKG samples, a low-pass filtering function is performed to remove high-frequency white noise by utilizing a digital finite impulse response filter. Finally, the heart rate is determined from the previously filtered EKG samples by a peak detection technique that performs a recursive algorithm to determine both the amplitude of the remaining peaks in the EKG samples and the period between these peaks.

Heart rate detection utilizing autoregressive analysis

Disclosed is a digital arrangement utilizing linear predictive coding for equalizing over a desired frequency spectrum the variable attenuation of a voice-frequency message signal transmitted on a communication line. The arrangement comprises a digital signal processor, program memories for storing program instruction sets, and a data memory for storing samples of a spectrally white test signal that has been variably attenuated by the line. Under control of one instruction set that incorporates linear predictive coding, the processor uses the stored test samples to calculate the reflection coefficients of the line that characterize the variable attenuation of a signal on the line. Under the control of the other instruction set, the processor functions as a digital inverse filter employing the calculated reflection coefficients for equalizing over the desired frequency spectrum the variable attenuation of a voice-frequency message signal transmitted on the line.

Arrangement for equalizing the variable attenuation of a signal on a

Speech signal pitch detection uses the residual signal output of an LPC (linear prediction coder) filter. The residual signal (or its Hilbert transform) is divided into frames of 20 milliseconds, and each frame searched for signal peak periodicity by first detecting the highest peak, then detecting a set of equally-spaced selected samples from which the pitch period is determined. If no periodicity is found over three frames, an unvoiced decision is made.

Growing community of inventors

Park Ridge, IL, United States of America

Dimitrios P Prezas