摘要:

Annual audiometric testing mandated by Federal Regulations is creating an enormous, fragmented database. Epidemiologic studies, based on this information, can potentially elucidate important parameters related to the development of noise‐induced hearing loss. This paper illustrates the application of survival data analysis to the audiometric data of approximately 2500 people taken over as many as ten successive years. Survival analysis includes both the number of people who fail during a study (develop a standard threshold shift in this instance) as well as those who survive or are lost to follow‐up. In this case, we found that the survival function (the probability of surviving past a time interval) and the hazard function (the probability of surviving through a time interval given that the worker has survived to the beginning of the interval) are insensitive to the exposures at the facility. However, when these functions are stratified on age at first test, older workers are at higher risk than younger workers.

ISSN:0001-4966    

DOI:10.1121/1.2023605

出版商:Acoustical Society of America    

年代:1986

数据来源: AIP

摘要:

Normal‐hearing listeners with hearing losses simulated by masking noise benefit from wideband amplitude compression under conditions (e.g. no interference, most‐comfortable presentation level) for which listeners with sensorineural hearing impairments do not. Such findings suggest that speech intelligibility may be limited by degraded spectral and temporal resolution as well as reduced dynamic range for hearing‐impaired listeners. To combat the combination of reduced dynamic range and reduced frequency selectivity, we have developed a speech‐processing system which provides both widehand compression and a sharpening of short‐term spectral shape. The sharpening is accomplished by an orthogonal decomposition of the critical‐band spectrum (similar to a principal‐component decomposition) followed by expansion of the component weights which specify the peak‐valley structure of the spectrum. We compare speech intelligibility with this system to that with linear amplification and wideband compression alone for hearing‐impaired listeners. Preliminary results for CVC nonsense syllables indicate a modest benefit for the compression/sharpening system for a male speaker, but little benefit for a female speaker. [Work supported by NIH.]

ISSN:0001-4966    

DOI:10.1121/1.2023659

出版商:Acoustical Society of America    

年代:1986

数据来源: AIP

摘要:

A digital master hearing aid configured around a programmable digital filter was used to simulate hearing aids prescribed by four different procedures. The prescriptive procedures were modeled on those developed by Skinner, Byrne and Toonison, Lybarger, and Levittet al.Real‐time simulation allowed for adaptive paired‐comparison testing to be used in evaluating the prescribed hearing aids. Data were obtained for three signal levels corresponding to input levels of 60 and 70 dB SPL and an output level corresponding to the subject's most comfortable loudness level. Speech discrimination scores were obtained for each simulated hearing aid at each signal level, as well as paired‐comparison judgments of relative intelligibility and overall quality. No single prescriptive procedure was found to be uniformly superior to the other procedures considered. Differences between procedures were largest at low signal levels and smallest at the subject's most comfortable level. [Research supported by NIHR.]

ISSN:0001-4966    

DOI:10.1121/1.2023661

出版商:Acoustical Society of America    

年代:1986

数据来源: AIP

摘要:

Recent work has led to a new spectral model for pressure fluctuations at a rigid plane wall bounding a turbulent boundary layer, based on an explicit representation of structures in the boundary layer near the wall—bursts and sweeps—that exchange momentum within the layer and support the shear stress there [J. M. Witting, Noise Control. Eng. J.26, 28—43 (1986)]. The fluctuating wall pressure results from the contributions of a collection of independent bursts and sweeps. Each burst and sweep is modeled as a dipole flow that moves with the local mean flow, has a finite duration, and has the correct strength to mix the fluid through a Prandtl mixing length above and below its center. This paper describes some of the salient features and extensions of the model, emphasizing (1) the physical interpretation of the model structure and its three adjustable parameters, which are sufficient to yield predictions for all frequencies and wavelengths, (2) the ranges of frequencies and wavenumbers over which the model can be applied with confidence, based both on theory and past experiments, and (3) the applicability of the modeling approach to other situations.

ISSN:0001-4966    

DOI:10.1121/1.2023667

出版商:Acoustical Society of America    

年代:1986

数据来源: AIP

摘要:

A detailed analytical model of ring‐stiffened cylindrical shell vibration is presented. The model is to accept excitation due to turbulent boundary layer pressure fluctuations. The mathematical model is capable of predicting the acoustical effects of a compliant decoupling layer attached to the external surface of the shell structure. Both the shell vibration levels and farfield acoustic pressures are predicted. A few focused example results are presented with and without acoustical decouplers.

ISSN:0001-4966    

DOI:10.1121/1.2023670

出版商:Acoustical Society of America    

年代:1986

数据来源: AIP

摘要:

Plumbing noise measurements on valves used for domestic lavatories, showers, tubs, and toilets have been measured utilizing the approach outlined in ISO Standard 3822. This technique evaluates the flow induced structural vibration transmitted into the piping system. The results clearly indicate that the principal source of noise is turbulence within the valve mechanism and not due to excessive velocities in the piping. Comparisons will be provided between American and European products. Several standard methods of isolating pipe wall vibration from building components are also evaluated.

ISSN:0001-4966    

DOI:10.1121/1.2023675

出版商:Acoustical Society of America    

年代:1986

数据来源: AIP

摘要:

Articulatory phonetics describes speech as a sequence of overlapping articulatory gestures, each of which may be associated with a characteristic ideal target spectrum. In normal speech, the idealized target gestures for each speech sound are often never attained, and the speech signal exhibits only transitions between such (implicit) targets. It has been suggested that the underlying speech sounds can only be recovered by reference to detailed knowledge of the gestures by which individual speech sounds are produced. It will be shown that it is possible to decompose the speech signal into overlapping “temporal transition functions” using techniques which make no assumptions about the phonetic structure of the signal or the articulatory constraints used in speech production. Previous work has shown that these techniques can produce a large reduction in the information rate needed to represent the spectral information in speech signals [B.S. Atal, Proc. ICASSP83, 2.6, 81–84 (1983)]. It will be shown that these methods are able to derive speech components of low bandwiths that vary on a time scale closely related to traditional phonetic events. Implications for perception and the application of such techniques both for speech coding and as a possible front end for speech recognition will be discussed.

ISSN:0001-4966    

DOI:10.1121/1.2023680

出版商:Acoustical Society of America    

年代:1986

数据来源: AIP

摘要:

The integration of grammatical with acoustical knowledge sources in the BBN continuous speech recognition system, BYBLOS, and the resulting effects on performance are described. The system consists of feature extraction, acoustical scoring, and linguistic scoring. Feature extraction is based on vector quantized reel‐warped cepstral coefficients. Acoustical scoring is derived from a hidden Markov model for each word, where word models are based on phonetic spellings so that models can be computed for words that have never been trained. The linguistic model is represented as a finite automaton derived automatically from a context‐free specification of the task‐domain syntax and semantics. It is shown how recognition performance varies with properties of the grammars. Word recognition accuracies of over 98% have been achieved in continuous speaker‐dependent mode for 350‐word tasks with grammars having maximum perplexity in the range of 20 to 60. [Work supported by DARPA and monitored by NAVELEX.]

ISSN:0001-4966    

DOI:10.1121/1.2023686

出版商:Acoustical Society of America    

年代:1986

数据来源: AIP

摘要:

The goal is to develop a speaker‐independent, semivowel detector‐classifier for use with continuous speech. Preliminary analysis of semivowels within continuous speech indicates large within‐(and between) speaker variability in the respective semivowel formant frequencies, amplitudes, and durations of steady states and transitions. Deviations from textbook descriptions based on isolated words are significant. Coarticulation effects are also large. Nevertheless, through the use of energy, energy ratios, zero crossings, and signal envelopes, a logical analytical scheme was developed that classifies segments of continuous speech as: semivowel, high‐front vowel, vowel, nasal, fricative, stop, or plosive. The specific characteristics of this scheme will be discussed and the computer‐generated displays that were used will be presented. Results of using this scheme on a large corpus of phonetically labeled speech will be presented. [Work supported, in part, by DARPA.]

ISSN:0001-4966    

DOI:10.1121/1.2023689

出版商:Acoustical Society of America    

年代:1986

数据来源: AIP

摘要:

A standard coherence measure has been used to give an estimate of signal‐to‐noise ratio (certainty indicators) as a function of frequency. The two spectra which are used for this measure are the noisy speech spectrum and a model spectrum derived from pitch period estimates of the noisy speech. It is intended that the certainty indicators be used as either weights for matches in automatic recognition or as multiplicative weights for reconstruction of speech with improved intelligibility. Two cases of noisy speech have been studied: single words contaminated by bandpass‐filtered white noise (S1) and sentences contaminated by unfiltered white noise (S2). For case S1 it has been found that the certainty indicators will pinpoint the noisy band for voiced speech and will, as expected, not reliably indicate the noisy band for unvoiced speech. For case S2, preliminary results indicate that the quality of the enhanced speech has dropped, yet the intelligibility is slightly improved. [Work supported by NSF and Boeing Computer Services.]

ISSN:0001-4966    

DOI:10.1121/1.2023696

出版商:Acoustical Society of America    

年代:1986

数据来源: AIP