Consistency among speech parameter vectors: application to predicting speech intelligibility.

M H Power; L D Braida

doi:10.1121/1.417243

Home/Archive/PubMed

PubMed · 8969488

Consistency among speech parameter vectors: application to predicting speech intelligibility.

M H Power·L D Braida

View original source ↗Export metadata ↓

Abstract

Previous researchers interested in physical assessment of speech intelligibility have largely based their predictions on preservation of spectral shape. A new approach is presented in which intelligibility is predicted to be preserved only if a transformation modifies relevant speech parameters in a consistent manner. In particular, the parameters from each short-time interval are described by one of a finite number of symbols formed by quantizing the output of an auditory model, and preservation of intelligibility is modeled as the extent to which a one-to-one correspondence exists between the symbols of the input to the transformation, and those of the output. In this paper, a consistency-measurement system is designed and applied to prediction of intelligibility of linearly filtered speech and speech degraded by additive noise. Results were obtained for two parameter sets: one consisting of band-energy values, and the other based on the ensemble interval histogram (EIH) model. Predictions within a class of transformation varied monotonically with the amount of degradation. Across classes of transformation, the predicted effect of additive-noise transformations was more severe than typical perceptual effects. With respect to the goal of achieving predictions that varied monotonically with human speech-perception scores, performance was slightly better with the EIH parameter set.

Explore related subjects

Auditory Threshold Humans Noise Perceptual Masking Phonetics Speech Perception Journal Article Research Support, U.S. Gov't, P.H.S.

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

M H Power, L D Braida. 1996. Consistency among speech parameter vectors: application to predicting speech intelligibility.. https://doi.org/10.1121/1.417243

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2007

A dual-process integrator-resonator model of the electrically stimulated human auditory nerve.

A phenomenological dual-process model of the electrically stimulated human auditory nerve is presented and compared to threshold and loudness data from cochlear implant users. The auditory nerve is modeled as two parallel processes derived from linearized equations of conductance-based models. The first process is an integrator, which dominates stimulation for short-phase duration biphasic pulses and high-frequency sinusoidal stimuli. It has a relatively short time constant (0.094 ms) arising from the passive properties of the membrane. The second process is a resonator, which induces nonmonotonic functions of threshold vs frequency with minima around 80 Hz. The ion channel responsible for this trend has a relatively large relaxation time constant of about 1 ms. Membrane noise is modeled as a Gaussian noise, and loudness sensation is assumed to relate to the probability of firing of a neuron during a 20-ms rectangular window. Experimental psychophysical results obtained in seven previously published studies can be interpreted with this model. The model also provides a physiologically based account of the nonmonotonic threshold vs frequency functions observed in biphasic and sinusoidal stimulation, the large threshold decrease obtained with biphasic pulses having a relatively long inter-phase gap and the effects of asymmetric pulses.

Auditory Threshold↗

PubMed2007

Connexin 26 deafness is not always congenital.

OBJECTIVE: Hearing loss associated with mutations of the GJB2, the gene encoding Connexin 26 (Cx26), is described as a prelingual, bilateral, prevalently stable sensorineural defect ranging in severity from mild to profound. Despite many clinical studies, there is still a limited knowledge about the severity of Cx26 hearing loss at birth, in the postnatal period or in early infancy; some authors have reported about a possible variable age of onset. The aim of this work was to investigate the characteristics of Cx26 hearing loss and test the hypothesis of a postnatal sudden and severe deterioration of the hearing capacity in cases with uncertain age at onset. METHODS: We have studied 79 children with molecularly documented biallelic Cx26 hearing loss by evaluating longitudinal audiometric characteristics and the results of a questionnaire administered to the parents, regarding the auditory behavior of their children at 3 and 6 months of age. RESULTS: More than 50% of children with profound hearing loss were described as having normal auditory behavior at three months of age, and at least 20% of these children were consistently reported by the parents to maintain normal auditory development up to 6 months of age. None of the studied children showed significant progression in hearing loss from the time of diagnosis through their last follow-up. In a few cases these reports were supported by objective audiometric evaluations. CONCLUSIONS: Based on these data, we hypothesize that Cx26 profound hearing loss may be not always congenital, with the possibility of an early window of "functional time" before the final defect is established. The hypothesis of an early auditory input makes timing of detection and intervention critical to minimize the deleterious effects of a functional deprivation. In a near future, the combination of genetic testing with universal neonatal hearing screening and audiological surveillance will provide invaluable information about the natural history of the most frequent sensory defect in infancy and will consequently allow to maximize the quality of intervention.

Auditory Threshold↗

PubMed2006

Parental suspicion of hearing loss in children with otitis media with effusion.

OBJECTIVE: This study aimed to evaluate the parental suspicion of hearing loss in children with otitis media with effusion (OME). As part of a population-based survey in a screening programme among 6- to 7-year-old Chinese children in Hong Kong, OME cases and controls were studied for the value of parental observations in the prediction of OME and hearing test results. SUBJECTS AND METHODS: Prior to the otoscopic and tympanometric examination of the children on school premises, a self-administered binary-choice question was sent to the parents asking whether there was any suspicion of hearing impairment. Positive screens and randomly selected negative screens were seen in a hospital clinic for the confirmation of case and control status 2 to 3 weeks after the school screening. Aural examination under microscopy, repeated tympanometry and stapedial reflex testing, and pure-tone audiometry (PTA) were conducted, and 117 cases and 159 controls were included in this study. RESULTS: The average PTA conductive threshold levels in the individual children with OME ranged from 3.8 dB to 40.0 dB with a group mean of 17.0 dB in the better-hearing ears. Parental suspicion of hearing deficit was significantly associated with OME (p<0.001) but not PTA findings (p=0.686). The sensitivity of parent-suspected hearing impairment to detect OME however was very low (19.7%). DISCUSSION: In other words, if we had relied on parental suspicion as the first screening, at least 80% of the OME cases would have been missed. We conclude that the parental suspicion of hearing loss is inadequate for the identification of mild hearing loss as caused by OME. Health education is recommended to improve parental awareness of the disease.

Auditory Threshold↗