CV Context Research Articles

An Electropalatographic Study of Korean Fricatives /s/ and /s*/

Objectives: The present study is aimed an electropalatographic investigation of Korean fricatives /s/ and /s*/. All data from this study is based on the production of 5- adult speakers who speak Korean as their first language.Methods: Electropalatography (EPG) is an instrumental technique for showing tongue and palate contact during speech. The participants recorded /s/ and /s*/ in CV (/si/, /sɑ/, /su/, /s*i/, /s*ɑ/, /s*u/) and VCV (/ɑsi/, /ɑsɑ/, /ɑsu/, /ɑs*i/, /ɑs*ɑ/, /ɑs*u/) context using EPG and then, four variables were analyzed: row of medial groove (RMG), medial groove width (MGW), medial groove length (MGL), and maximum contact frame (MCF).Results: There was significantly greater medial groove width and maximum contact frame for / s*/ compared to /s/. The difference between /si/ and /sɑ/ was also found in the row of medial groove, medial groove width and maximum contact frame. /si/ was produced further back, and had a longer medial groove length. Also, it showed more total tongue-palate contact than /sa/.Conclusion: A various range of palatal contact patterns were found between typical adults for perceptually normal /s/ and /s*/. The results provided the information about tongue–palate pattern of Korean fricatives /s/ and /s*/. These findings would be utilized to build the basic data and useful information for Korean EPG system.

Investigating the holographic complexity in Einsteinian cubic gravity

In this paper, we investigate the holographic complexity of a small mass AdS black hole in Einsteinian cubic gravity by using the “complexity equals action” (CA) and “complexity equals volume” (CV) conjectures. In the CA context, the late-time growth rate satisfies the Lloyd bound for the k=0 and k=1 cases but it violates it for the k=-1 case in the first-order approximation of the small mass parameter. However, by a full-time analysis, we find that this late-time limit is approached from above, which implies that this bound in all of these cases will be violated. In the CV context, we considered both the original and the generalized CV conjectures. Differing from the CA conjecture, the late-time rate here is non-vanishing in the zeroth-order approximation, and this shows that the Lloyd bound is exactly violated even in the late-time limit. These results show numerous differences from the neutral case of the Einstein gravity in both the CA and the CV holographic contexts where all of their late-time results saturate the Lloyd bound. These differences illustrate the influence of the higher curvature correction in Einstein gravity.

Perception of /ra/-/la/ contrast in different contexts: mono-syllable vs. sentence

A strict assumption that underlies categorical perception hypotheses is that two speech sounds are discriminable only when they cross a categorical boundary emerging from an identification function. A number of researchers have attempted to show the categorical perception of the sounds in question (particularly consonants) through discrimination and identification tests. However, these tests are usually used for mono-syllables or mono-syllabic words. In this study, we first investigated whether the perception of /ra/-/la/ contrast indicate any categorical perception in a mono-syllabic CV context, using synthesized syllables. Next, we tested whether categorical perception can be also observed in a sentence through perceptual discrimination and identification experiments. Results showed that (1) a discrimination peak predicted by the identification function was obtained only for the mono-syllabic context, and (2) discrimination accuracy in the sentence condition was consistently low. These results suggest that categorical perception in a strict sense may not be evident in the perception of a sentence.

Neural representation of speech sounds: Study using frequency following response.

The neural encoding of an acoustic signal begins in the auditory nerve and travels to the auditory brainstem and further to the auditory cortex. Previous studies have used nonspeech signals like tones and clicks to evaluate the integrity and synchrony of the auditory pathway. Most of the previous research has focused on how any acoustic signal is perceived by using behavioral methods. One of the main areas in understanding how we hear the signal depends on how this acoustic signal is represented in the auditory pathway. A number of studies have studied the acoustic parameters that influence the perception of speech, but very few of them have focused on how these acoustic changes are represented in the auditory pathway. Since the signal representation in the auditory pathway is very crucial to how the signal is perceived, studies that focus on the relationship between the changes in the input acoustics and its influence on neural representation become very essential. In the current project the neural representation of isolated vowels and vowels in CV context was recorded using human frequency following response. The results will be discussed in the paper.

Use of Linear Frequency Transposition in Simulated Hearing Loss

To demonstrate the usability of new acoustic cues provided by a commercial hearing aid using linear frequency transposition. The importance of auditory training in realizing the benefit was evaluated. Quasi-experimental study design. All subjects were tested both with conventional amplification and with linear frequency transposition before and after directed training. A total of nine normal hearing participants with simulated high frequency hearing loss at and above 1600 Hz participated in the study. All subjects were native English speakers ranging in age from 18 to 24 years. Identification of voiceless phonemes in CV, VC, and CVC context processed with and without linear frequency transposition was evaluated. Identification test was carried out four times. Between test trials, participants completed 15 minutes of self-paced training using transposed stimuli. Prior to any training, transposition did not improve phoneme identification scores. Training of 30 minutes improved the overall identification scores of the transposed stimuli over the nontransposed stimuli by 14.4%. The results demonstrated that frequency transposition produces acoustic cues that normal hearing listeners with a simulated hearing loss at and above 1600 Hz may be trained to utilize.

Perceptual overshoot in listeners with cochlear implants

Perceptual overshoot (PO), a phenomenon in which the boundary for perceived vowel categories shifts as a result of preceding formant transitions, has been demonstrated previously with synthetic vowels in CV contexts in normal-hearing individuals. In one of the previous studies by the author, the same phenomenon was tested with sinewave analogs that mimicked the synthetic vowels. The results demonstrated that PO could be elicited for sinewave analogs after training on categorization of sinewave steady states. These findings suggest that PO may be partly mediated by general processes in the auditory system. In the current study, subjects with cochlear implants were presented with synthetic vowel and sinewave continua with and without transitions. It was predicted that PO would be greatly diminished or absent for this population because of the degraded nature of the spectral input. Implications of the results for theories of speech perception will be presented.

Perceptual overshoot with speech and nonspeech sounds

One of the basic quests in speech perception research has been to find the differences or similarities in the mechanisms involved in the perception of speech and nonspeech sounds. The current study will address the differences in perception of speech and nonspeech signals by comparing the perceptual overshoot in synthetic vowels and sinewave acoustic replicas of the synthetic vowels. Lindblom and Studdert‐Kennedy (1967) demonstrated that the perceptual boundary for steady state vowels and that for vowels in a CV context with F2 transition are different. They called this phenomenon a perceptual compensation or perceptual overshoot. In the current study the perceptual boundaries for synthetic steady state vowels, steady state sinewave acoustic replicas of vowels, a vowel in the CV context with F2 transition and sinewave acoustic replicas of vowels in the CV context are compared. The results will be discussed in the poster. For Speech Communication Best Student Paper Award.

Speech-sound perception in normal and learning-disabled children: Effect of lengthened CV transition duration

Previous research has established that many children with learning problems exhibit perceptual deficits in response to certain auditory stimuli, such as stop consonants in CV context. In order to investigate the precise acoustic features of stop consonants that pose perceptual difficulties for these children, discrimination thresholds were compared for normal and learning-disabled children along two separate synthetic /da/–/ga/ continua that differed only in the duration of the formant transitions. Results showed that simply lengthening the formant transition duration from 40 to 80 ms did not result in improved discrimination thresholds for the learning-disabled children. Similarly, an electrophysiologic response that is known to reflect the brain’s preconscious ability to detect a change from one auditory stimulus to another indicated diminished responses in these children to /da/ versus /ga/, regardless of transition duration. This finding suggests that the brevity of the CV formant transitions is not the sole acoustic basis for the observed perceptual deficits. Furthermore, signal processing techniques aimed at enhancing perception of stop consonants likely need to consider other acoustic features, such as relative amplitude and frequency characteristics, instead of (or in addition to) CV transition duration. [Work supported by NIH R01 DC 01510.]

Acoustic parameters for place of articulation identification and classification of Spanish unvoiced stops

The analysis of the acoustic parameters which best summarize the cues to phone discrimination for the language under consideration should be a previous step in acoustic-phonetic decoding, regardless of the methodology to be used. The Spanish language has not been widely analyzed from this point of view. This work deals with the acoustic discrimination of Spanish stop consonants. Our main goal was to find a reliable and reduced set of parameters for place of articulation identification of Spanish unvoiced stops. On the basis of the obtained parameters, two automatic classifiers were developed and tested. Only the acoustic features of the burst segment, automatically segmented from the speech waveform, were considered in the parameter estimation. The analysis of these features was carried out in both the time and frequency domains over a CV context corpus uttered by 6 speakers. In the first case, the classifier was designed as a procedural form. Alternatively, in the second case a statistical classifier was obtained from a previous automatic discriminant analysis of the parameters. Both classifiers were tested over a CV context corpus uttered by 40 new speakers not included in the analysis corpus, which resulted in a good rate of identification.

Vowel discrimination in a phonologically neutralized context.

The perception of vowel quality is not only affected by the formant structure of the vowel, but by the nature of the surrounding phonetic context. For example, it has been demonstrated [R. A. Fox, in Speech Perception, Production and Linguistic Structure, edited by Y. Tohkura et al. (IOS Press, 1992)] that listeners’ ability to make reliable vowel quality distinctions are lowered when the vowels are placed in a context that introduces phonological neutralization (e.g., post‐vocalic [■]). One possible explanation for this phenomenon is that of listener bias, which should be eliminated using the appropriate signal detection methodology. The present study describes the results of an experiment that obtained d’ measurements of one‐step differences in an [i]‐[■] continuum in CV, CVr, CVl, and CVd contexts under fixed‐level, minimum uncertainty conditions. Results demonstrate after several days experience with the stimuli and methodology that the discrimination differences as a function of context were eliminated. However, a post‐training identification test still showed context effects on vowel categorization. These data will be discussed in terms of the possible nature of this listener bias in vowel perception.

Analysis of nasal consonants using perceptual linear prediction.

Until recently, speech analysis techniques have been built around the all-pole linear predictive model. This study examines the effectiveness of using the perceptual linear predictive method for analyzing nasal consonants. Six speakers (three men and three women) produced 300 CV syllables with initial nasal consonants /m/ and /n/. A threshold-based boundary detection algorithm was developed to extract nasal segments from the CV contexts. Poles of a fifth-order perceptual linear predictive model were calculated and the frequency of the second pole was used to characterize the place of articulation of nasal consonants. Results indicated that the frequency for the second transformed pole was significantly lower for /m/ than for /n/ and was independent of factors such as a vowel context and gender of the speaker. A nasal identification rate of 86% was obtained based on the frequency of the second pole. The use of the perceptual linear predictive method may thus overcome some difficulties associated with analyzing nasal consonants.

Automatic identification of place of articulation of nasal consonants using the PLP method

Until recently, techniques on speech analysis have been built around the all‐pole linear predictive model. This study demonstrates the advantage and effectiveness of using the perceptual linear predictive (PLP) method in analyzing nasal consonants. Six speakers (three men and three women) produced 300 CV syllables with initial nasal consonants /m/ and /n/. A threshold‐based boundary detection algorithm was developed to extract nasal segments from CV contexts. Poles of the fifth‐order PLP model were calculated to characterize the place of articulation of nasal consonants. Results indicated that the frequencies for the transformed poles, particularly for the second pole, were significantly lower for /m/ than for /n/ and were independent of factors such as vowel context and gender of the speaker. A nasal identification rate of 86% was obtained based on the frequency of the second pole. The use of the PLP method clearly has overcome difficulties associated with the antiresonance in analyzing nasal consonants.

Preliminary evaluation of a synthetic Tadoma system

At the 105th ASA meeting we presented results of our research on the Tadoma method of tactile speech reception and described our ongoing development of a “synthetic Tadoma system” for use as a research tool in studying transformations of Tadoma [Reed et al., J. Acoust. Soc. Am. Suppl. 1 73, S26 (1983)]. In the synthetic Tadoma system, signals recorded from a talker's face are used to control an artificial mechanical face. Preliminary evaluation of the system has involved roving ABX discrimination tests among sets of 15 vowels in CVC contexts and among sets of 24 consonants in CV contexts. Two naive adult subjects, with simulated deafness and blindness, have been tested. Average results for vowel discrimination (68%) were equal to those obtained in a comparable study of natural Tadoma [Reed et al., J. Speech Hear. Res. 21, 625–637 (1978)]. Results for consonant discrimination were lower for the synthetic system (68% vs 78%), but were improved (to 74%) following certain modifications to the system. [Work supported by NSF.]

Stop-vowel coarticulation in 3-year-old, 5-year-old, and adult speakers

The perturbation of second formant onset values in CV contexts (C=/b,d,g/) caused by the vowels /i/ and /u/ was measured in CVC tokens spoken by three 3-year olds, 5-year olds, and adults. These preliminary results suggest similar patterns of stop-vowel coproduction in all three groups.

Coarticulation in sequences of two nonhomorganic stop consonants: perceptual and acoustic evidence.

This study investigated whether any perceptually useful coarticulatory information is carried by the release burst of the first of two successive, nonhomorganic stop consonants. The CV portions of natural VCCV utterances were replaced with matched synthetic stimuli from a continuum spanning the three places of stop articulation. There was a sizable effect of coarticulatory cues in the natural-speech portion on the perception of the second stop consonant. Moreover, when the natural VC portions including the final release burst were presented in isolation, listeners were significantly better than chance at guessing the identity of the following, "missing" syllable-initial stop. The hypothesis that the release burst of a syllable-final stop contains significant coarticulatory information about the place of articulation of a following, nonhomorganic stop was further confirmed in acoustic analyses which revealed significant effects of CV context on the spectral properties of the release bursts. The relationship between acoustic stimulus properties and listeners' perceptual responses was not straightforward, however.

Airflow characteristics of fricative consonants produced by normally hearing and hearing-impaired speakers.

The purpose of this study was to investigate the rates of airflow for fricative consonants produced by normally hearing speakers and by hearing-impaired speakers who varied in degree of hearing loss and speech intelligibility. Subjects consisted of 20 normally hearing adults (10 men and 10 women), 20 moderately-to-severely hearing-impaired adults with intelligible speech (10 men and 10 women), and 20 profoundly hearing-impaired adults with semi-intelligible speech (10 men and 10 women). Average airflow rates were obtained for the fricatives /s/, /z/, /f/, and /v/ in CV and VCV syllables. In general, it was found that normally hearing, intelligible hearing-impaired, and some semi-intelligible hearing-impaired speakers produced voiceless fricatives with substantially greater airflow rates than they produced voiced cognates. The data also revealed that both male and female semi-intelligible hearing-impaired speakers produced fricatives in the CV context with greater amounts of airflow and fricatives in the VCV environment with less airflow than did the normally hearing and intelligible hearing-impaired speakers. Further, the male and female semi-intelligible hearing-impaired speakers failed to demonstrate the trend, as exhibited by the other speakers, for lingua-alveolar fricatives to be produced with greater air flow than the labiodental fricatives. The data from this investigation provide additional evidence to support the assumption that some hearing-impaired speakers fail to control and coordinate the biomechanical aspects of speech properly.

Vibrotactile perception of phonetic features of speech: A comparison of single‐channel and multi‐channel aids

In an earlier paper [C. A. Beachler and A. E. Carney, J. Acoust. Soc. Am. Suppl. 1 69, S123 (1981)], two vibrotactile devices, one single‐channel and one multi‐channel, were compared with regard to their ability to transmit suprasegmental aspects of speech. Subjects trained with a single‐channel device were significantly better trained than those trained with a multi‐channel device in the identification of number of syllables in a word, syllable stress and intonation contour. As a follow up, two additional groups of subjects were trained with the same two different vibrotactile devices to recognize 20 consonants and eight vowels, all in a CV context. Subjects were trained in three conditions—visual only, vibrotactile only, and visual plus vibrotactile. Results indicated that there was essentially no difference between instruments in the percentage of phonemes transmitted correctly. Some slight differences between instruments were observed in the pattern of errors and the particular phonemes which were transmitted best by each instrument. These data are discussed with regard to the selective applicability of these devices to speech training for the hearing impaired.

Perceptual assessment of coarticulation in sequences of two stop consonants

This study investigated whether any perceptually useful coarticulatory information is carried by the release bursts and formant transitions of two successive, nonhomorganic stop consonants. The VC or CV portions of natural VCCV utterances were replaced with matched synthetic stimuli from a VC or CV continuum spanning the three places of stop articulation. When the VC and CV portions in the resulting hybrid VCCV stimuli were separated by a fixed silent interval, the context in which the natural portion has been produced had no influence on listeners' identification of the synthetic portion, suggesting that VC and CV formant transitions and CV release bursts contained no perceptually salient coarticulatory cues. However, when a natural VC portion was separated from a synthetic CV portion by the original closure interval, which included a brief release burst of the first stop, there was a sizeable effect of the original CV context on the perception of the second stop consonant. Thus the release burst of a sy...

Some aerodynamic characteristics of plosive consonants produced by hearing-impaired speakers.

The purpose of this study was to investigate the rate of oral air flow for plosive consonants produced by 10 normal hearing adult males, 10 hearing-impaired adult males who demonstrated intelligible speech, and 10 hearing-impaired adult males with semi-intelligible speech. The peak rate of oral air flow was measured during production of CV and VCV syllables consisting of the vowels /a/ and /i/ and the consonants /b/, /p/, /d/, and /t/. In general, the results indicated that a significant voiced/voiceless distinction, for rate of air flow, occurred for the normally hearing and intelligible hearing-impaired subjects but not for the semi-intelligible subjects. In addition, the semi-intelligible hearing impaired, in the CV context, produced voiced plosives with significantly greater air flow and voiceless plosives with significantly less air flow; and, produced both voiced and voiceless plosives, in the VCV context, with significantly less air flow, when compared with the normally hearing and intelligible hearing-impaired speakers.

Acoustic characteristics of correctly produced vowels in deaf children's speech

An analysis was performed on the acoustic characteristics of the vowels /i, æ, u/ in several CV contexts in the speech of deaf and normal hearing children. The CV syllables were embedded in sentences, and only those syllables judged by a group of phoneticians to have been correctly produced were analyzed for the deaf children. For all three vowels, there was greater variability among the deaf subjects than among the normal hearing subjects in the frequency location of the first two formants. The average differences in the F1 − F2 plots between the two groups of speakers varied systematically depending upon the vowel produced and the phonetic context in which it occurred. Differences between the two groups of speakers was observed for the F1/F2 ratio as well as for the absolute frequency location of the formants. These data indicate that vowels in deaf children's speech can be correctly identified even though the absolute and relative formant values differ from that of normal hearing speakers. [Research supported by PHS Grant ♯09252 from NINCDS.]