Gregory C. Burnett

Speech articulator measurements using low power EM-wave sensors.

Very low power electromagnetic (EM) wave sensors are being used to measure speech articulator motions as speech is produced. Glottal tissue oscillations, jaw, tongue, soft palate, and other organs have been measured. Previously, microwave imaging (e.g., using radar sensors) appears not to have been considered for such monitoring. Glottal tissue movements detected by radar sensors correlate well with those obtained by established laboratory techniques, and have used to estimate a voiced excitation function for speech processing applications. The noninvasive access, coupled with the small size, low power, and high resolution of these new sensors, permit promising research and development applications in speech production, communication disorders, speech recognition and related topics.

Comparison between electroglottography and electromagnetic glottography

Newly developed glottographic sensors, utilizing high-frequency propagating electromagnetic waves, were compared to a well-established electroglottographic device. The comparison was made on four male subjects under different phonation conditions, including three levels of vocal fold adduction (normal, breathy, and pressed), three different registers (falsetto, chest, and fry), and two different pitches. Agreement between the sensors was always found for the glottal closure event, but for the general wave shape the agreement was better for falsetto and breathy voice than for pressed voice and vocal fry. Differences are attributed to the field patterns of the devices. Whereas the electroglottographic device can operate only in a conduction mode, the electromagnetic device can operate in either the forward scattering (diffraction) mode or in the backward scattering (reflection) mode. Results of our tests favor the diffraction mode because a more favorable angle imposed on receiving the scattered (reflected) signal did not improve the signal strength. Several observations are made on the uses of the electromagnetic sensors for operation without skin contact and possibly in an array configuration for improved spatial resolution within the glottis.

The use of glottal electromagnetic micropower sensors (GEMS) in determining a voiced excitation function

Recent experiments using a portable, extremely low‐power electromagnetic motion sensor to detect the motion of the posterior tracheal wall during speech production will be presented. The motion of the wall may be related to the driving subglottal pressure through a lumped element circuit model, leading to an approximation to the voiced excitation function of the human vocal tract. Using the excitation and the recorded spoken audio, a stable and accurate transfer function of the vocal tract may be calculated every few glottal cycles in near real‐time. The excitation function may be used to calculate very accurate pitch information at low cost, and the transfer functions may be employed as an additional feature vector to enhance the performance of a new class of speech recognizers and synthesizers. [Work supported by NSF and DOE.]

Using signal processing techniques to improve finite element modeling

The use of signal processing parameter identification techniques to systematically improve finite element (FE) models of extended structures will be presented. FE models have been used with success for many years to simulate the response of structures, and work has recently been undertaken at LLNL to use FE models to model extended structures such as the Oakland/San Francisco Bay Bridge. At present, there is no way to systematically improve the FE models using measured displacements of the structures. Standard signal processing parameter identification techniques adapted for large order models have proven useful in both refining FE models given measured displacement and acceleration data as well as indicating points of failure for structures after damaging events such as simulated earthquakes. These abilities could prove useful in improving current FE models as well as speeding repair time for damaged structures. The presentation will identify the strengths and weaknesses of specific parameter identificat...

Voiced excitation functions calculated from micropower impulse radar information

Efforts underway at the Lawrence Livermore National Laboratory to use newly designed micropower impulse radars (MIR) to measure in real time the excitation function of the vocal tract will be presented. Studies undertaken in collaboration with the University of California at Davis and the University of Iowa with high‐speed laryngoscopic cameras, electroglottographs, flow masks, and subglottal pressure transducers have solidified the relationship between the signal returned by the MIR and the voiced excitation function of the vocal tract. As a result, for the first time a transfer function of the vocal tract can be calculated in real time and with unprecedented clarity for voiced speech. This new capability could have significant implications for improvements in speech recognition and speech synthesis processing.

Background speaker noise removal using combined EM sensor/acoustic signals

Recently, very low‐power EM radarlike sensors have been used to measure the macro‐ and micro‐motions of human speech articulators as human speech is produced [see Holzrichter et al., J. Acoust. Soc. Am. 103, 622 (1998)]. These sensors can measure tracheal wall motions, associated with the air pressure build up and fall as the vocal folds open and close, leading to a voiced speech excitation function. In addition, they provide generalized motion measurements of vocal tract articulator gestures that lead to speech formation. For example, tongue, jaw, lips, velum, and pharynx motions have been measured as speech is produced. Since the EM sensor information is independent of acoustic air pressure waves, it is independent of the state of the acoustic background noise spectrum surrounding the speaker. By correlating the two streams of information together, from a microphone and (one or more) EM sensor signals, to characterize a speaker’s speech signal, much of the background speaker noise can be eliminated in r...

Comparison of conventional acoustic and MIR radar/acoustic processing of speech signals

Applications of the micropower impulse radar (MIR) to speech research at the Lawrence Livermore National Laboratory has produced potentially new methods of speech processing. They include the accurate calculation of vocal tract transfer functions, formant, and pitch analysis, and basic phoneme synthesis. These speech parameters have traditionally been in the realm of all‐pole LPC calculations. Related research using the MIR radar has supplied an increasingly accurate voiced excitation function, which makes possible transfer function calculations using both poles and zeros, yielding more accurate formant information and more natural sounding synthesis. This paper compares the newly obtained results with traditional LPC and cepstral approaches and demonstrates the improvements based on experimental data from several male and female subjects. The radar data also allow extremely accurate pitch tracking, which is simpler and more robust than that calculated by traditional means. This information can significan...

Speaker verification performance comparison based on traditional and electromagnetic sensor pitch extraction

This work compares the speaker verification performance between a traditional acoustic‐only pitch extraction to a new electromagnetic (EM) sensor based pitch approach system. The pitch estimation approach was developed at the Lawrence Livermore National Laboratory (LLNL) utilizing Glottal Electromagnetic Micropower Sensors (GEMS, also see http://speech.llnl.gov/). This work expands previous pitch detection work by Burnett et al. [IEEE Trans. Speech and Audio Processing (to be published)] to the specific application of speaker verification using dynamic time warping. Clearly, a distinct advantage of GEMS is its insensitivity to acoustic ambient noise. This work demonstrates the clear advantage of the GEMS pitch extraction to improve speaker verification error rates. Cases with added white noise and other speech noise were also examined to show the strengths of the GEMS sensor in these conditions. The EM sensor speaker verification process operated without change over signal‐to‐noise (SNR) conditions rangin...

Damage detection for enhanced evaluation of structures

Monumental structures consist of public and private buildings, bridges and other transportation structures, dams, key industrial and scientific buildings and many special structures. Most of these structures are key to our social, commercial and technical infrastructure. All are vulnerable to damage or failure when stressed by abrupt, large‐scale extreme events, the most common of which are earthquakes and wind. Forward reaching simulations, structural dynamic analysis techniques, innovative sensing, signal processing and remote communication form the core of this effort. Here we discuss our solutions to the structural damage detection problem. We take a model‐based signal processing approach. This means that we exploit prior knowledge from first principle models for the structure to construct signal processing algorithms. Assuming that the models are representative of the actual structure, the model based approach offers significant performance improvement over algorithms based on measured data alone. Here we discuss the processor performance on both simulated and experimental data of a five‐story structure.

Uses of micropower radars for speech coding and applications

It has recently become possible to measure the positions and motions of the human speech organs, as speech is being articulated, by using micropower radars in a noninvasive manner. Using these instruments the vocalized excitation function of human speech is measured and thereby the transfer function of each constant vocalized speech unit is obtained by deconvolving the output acoustic pressure from the input excitation function. In addition, the positions of the tongue, lips, jaw, velum, and glottal tissues are measured for each speech unit. Using these data, very descriptive feature vectors for each acoustic speech unit were able to be formed. It is believed that these new data, in conjunction with presently obtained acoustic data, will lead to more efficient speech coding, recognition, synthesis, telephony, and prosthesis.