Higher Speech Intelligibility Research Articles

Fast estimation of Speech Transmission Index using the Reverberation Time: Comparison between predictive equations for educational rooms of different sizes

Ensuring a high speech intelligibility is an essential issue for educational rooms, especially in relation to the effectiveness of the student learning process. The Reverberation Time and the Speech Transmission Index are objective parameters commonly used to assess the acoustic quality of the rooms, according to their intended use. Especially the Speech Transmission Index can give interesting information about the speech intelligibility. As with the Reverberation Time, it would be very important to have simple equations aimed to the prediction of Speech Transmission Index values during the early design stage. There are different studies in the literature that propose equations, obtained on empirical basis, for a fast estimation of the Speech Transmission Index as a function of the Reverberation Time. In this paper, a comparison of the most significant equations proposed in the literature was done using the Reverberation Time and Speech Transmission Index values obtained from a measurement campaign at the School of Engineering of the University of Pisa on a sample of 11 educational rooms. The analysis was carried out for educational rooms of different sizes, in order to highlight differences in the prediction accuracy depending on the volume of the rooms. From the analysis results it was possible to highlight the equation that shows the best accuracy in predicting the experimental data for the sample of investigated rooms and propose, through a regression analysis, two new equations for the rapid estimation of the Speech Transmission Index.

Using listening effort assessment in the acoustical design of rooms for speech

This study addresses the issue of an enhanced acoustical design of rooms for speech, which besides targeting high speech intelligibility also ensures minimal effort in speech reception. Speech-in-noise tests in the Italian language were proposed to normal-hearing young adults, both in situ, within an existing university classroom, and via headphones, using auralized signals obtained from acoustic simulations of the same environment. Later, auralization was used to investigate the effect of realistic modifications to the room acoustics (acoustical treatment of a wall, change of the room size) by altering the virtual model of the classroom. The speech reception performance was characterized by using both the number of words correctly recognized (speech intelligibility, IS) and two estimates of listening effort: the behavioral measure of response time (RT) and a subjective judgement on a rating scale (LE). Firstly, the correspondence between the IS, RT and LE results in situ and in auralized conditions was considered and discussed. Then, the effectiveness of the three metrics in outlining the effect of the acoustic changes of the room was analyzed. The results showed that there were no differences between the compared acoustic conditions in terms of IS. The effects of the characteristics of the room acoustics were instead discriminated when RT and LE were considered, with the greatest number of significant differences observed by using RT. Using RT therefore seems to be an effective and promising strategy to better discern the effects of the room acoustics and to enhance the acoustical design of rooms for speech.

Spatial Selectivity of Hearing in Speech Recognition in Speech-shaped Noise Environment

A variety of perceptual features can be used for the successful separation of information flows and higher speech intelligibility. The binaural system based on the spatial allocation of speech signals plays the most important role in attaining this goal. This review discusses how the mechanisms of spatial hearing provide selective attention to a target speech source and promote the recognition of a masked target signal.

The face of the facility

Lobbies and atria are a facility’s initial destination, the point of departure, the information center, and the security checkpoint. They define a building’s aesthetic, connect to the rest of the facility, require high speech intelligibility at key locations, and demand durability. The requirements can fluidly fluctuate throughout design, and the acoustics need to keep pace. This paper will discuss different projects where the lobby was central to a successful design. Highlights include a hospital atrium, wherein a balance was struck between speech intelligibility and speech privacy; an airport concourse where the prospect of tile grout lines was carefully considered; and a giant courthouse lobby with design goals at odds with design standards, and with bridges providing acoustical “shade” for security guards.

Corpus of deaf speech for acoustic and speech production research.

A corpus of recordings of deaf speech is introduced. Adults who were pre- or post-lingually deafened as well as those with normal hearing read standardized speech passages totaling 11 h of .wav recordings. Preliminary acoustic analyses are included to provide a glimpse of the kinds of analyses that can be conducted with this corpus of recordings. Long term average speech spectra as well as spectral moment analyses provide considerable insight into differences observed in the speech of talkers judged to have low, medium, or high speech intelligibility.

Intelligibility and Clarity of Reverberant Speech: Effects of Wide Dynamic Range Compression Release Time and Working Memory.

The purpose of this study was to examine the effects of varying wide dynamic range compression (WDRC) release time on intelligibility and clarity of reverberant speech. The study also considered the role of individual working memory. Thirty older listeners with mild to moderately-severe sloping sensorineural hearing loss participated. Individuals were divided into high and low working memory groups on the basis of the results of a reading span test. Participants listened binaurally to sentence stimuli simulated at a range of reverberation conditions and WDRC release times using a high compression ratio. Outcome measures included objective intelligibility and subjective clarity ratings. Speech intelligibility and clarity ratings both decreased as a function of reverberation. The low working memory group demonstrated a greater decrease in intelligibility with increasing amounts of reverberation than the high working memory group. Both groups, regardless of working memory, had higher speech intelligibility and clarity ratings with longer WDRC release times. WDRC release time had a larger effect on speech intelligibility under more reverberant conditions. Reverberation significantly affects speech intelligibility, particularly for individuals with lower working memory. In addition, longer release times in hearing aids may improve listener speech intelligibility and clarity in reverberant environments.

Simulation and overall comparative evaluation of performance between different techniques for high band feature extraction based on artificial bandwidth extension of speech over proposed global system for mobile full rate narrow band coder

This paper addresses a novel approach to investigate, study and simulate computation of high band (HB) feature extraction based on linear predictive coding (LPC) and mel frequency cepstral coefficient (MFCC) techniques. Further, HB features are embedded into encoded bitstream of proposed global system for mobile (GSM) full rate (FR) 06.10 coder using joint source coding and data hiding before being transmitted to receiving terminal. At receiver, HB features are extracted to reproduce HB portion of speech and for the same different extension of excitation techniques are applied and their results evaluated in terms of quality (intelligibility and naturalness) and bandwidth. MATLAB based e-test bench is created for implementing the proposed artificial bandwidth extension (ABE) coder following series of simulations, that are carried out to discover and gain insight about the performance of it using subjective [mean opinion score (MOS)] and objective [perceptual evaluation of speech quality (PESQ)] analysis. The results obtained for both the analyses advocate that proposed ABE coder outperforms proposed GSM FR NB (legacy GSM FR) coder. While the fact remains that, compared to LPC based parameterizations over ABE coder, MFCC parameterization results in higher speech intelligibility which is evident from obtained slightly better PESQ and MOS scores.

Impact of Background Noise and Sentence Complexity on Processing Demands during Sentence Comprehension.

Speech comprehension in adverse listening conditions can be effortful even when speech is fully intelligible. Acoustical distortions typically make speech comprehension more effortful, but effort also depends on linguistic aspects of the speech signal, such as its syntactic complexity. In the present study, pupil dilations, and subjective effort ratings were recorded in 20 normal-hearing participants while performing a sentence comprehension task. The sentences were either syntactically simple (subject-first sentence structure) or complex (object-first sentence structure) and were presented in two levels of background noise both corresponding to high intelligibility. A digit span and a reading span test were used to assess individual differences in the participants’ working memory capacity (WMC). The results showed that the subjectively rated effort was mostly affected by the noise level and less by syntactic complexity. Conversely, pupil dilations increased with syntactic complexity but only showed a small effect of the noise level. Participants with higher WMC showed increased pupil responses in the higher-level noise condition but rated sentence comprehension as being less effortful compared to participants with lower WMC. Overall, the results demonstrate that pupil dilations and subjectively rated effort represent different aspects of effort. Furthermore, the results indicate that effort can vary in situations with high speech intelligibility.

Parameters of speech intelligibility in patients with chronic sensoneural hearing loss depending on the hearing aid type

Nowadays audiologist has a large number of modern digital hearing aids (HA) to cover for different degrees of hearing loss. In some cases the same patient can be recommended several types and models of HA based on audiological data. Sometimes it creates difficulties in choice, especially for patients who have no experience of using the device. The objective of this study was to compare the results of the electroacoustic correction of various types of HA. The study covered 100 patients with similar audiological profile (2nd or 3rd degree of the chronic sensoneural hearing loss, sloping descendant or sloping audiogram). All the patients were divided into 4 equal groups depending on the type of device: behind­the­ear (BTE), BTE with slim tube (ВТЕ slim), receiver­in­the­canal (RIC) or complitely­in­the­canal (CIC). The HAs had identical acoustic characteristics. The speech intelligibility index (SII) was determined before using the hearing aid, after 12 weeks of hearing aid use without HA and with HA. Before HA using the average SII was 47.5%. 3 months after SII using without the HA was in average 49.7% with significant improvement in the RIC and CIC groups. SII with HA 3 months after HA using averaged 83.27% (81.73% in the BTE group, 80.94% in the BTEslim group, 85.47% in the RIC group and 84,94% in the CIC group). In this case SII with HA was significantly higher in the RIC and CIC groups. The obtained results prove higher speech intelligibility parameters in patients with the 2nd or 3rd degree of the chronic sensoneural hearing loss with RIC and CIC devices.

Proposals and Analysis of the Effects of Acoustic Corrections in a Modern Church

It is rather common experience to observe that many contemporary churches, in contrast with the effort to achieve refined architectonic shapes and finishes, exhibit poor acoustic quality, such as high reverberation times and low speech intelligibility. The causes of such deficiencies are ordinarily generated by the smooth and hard surfaces of walls, floor and ceiling that focus most of sound energy in late reflections, diminishing the effectiveness of direct and early reflected sound. Recently, the authors have been involved for analyzing the acoustic behavior of a modern church situated in Ragusa, (Italy), with the aim to individuate suitable interventions for improving the acoustic quality of that church. Preliminarily, a measurement survey was conducted to evaluate the main acoustic indices proposed by ISO 3382 (T30, STI, EDT, C80, D50) and portray the current acoustic climate. After that, the Authors developed a simulation on a 3D model of the church, in order to calibrate the model comparing measured and simulated acoustic parameters. The previously described activities are necessary for testing the reliability and accuracy of the prediction of the acoustic simulations. The subsequent step of the work consists in the choice of suitable interventions for the acoustic corrections; three different typologies of materials were evaluated. Globally, the acoustic corrections allow to obtain noteworthy improvements of the acoustic quality of the church, e.g. the T30 decrease from 7.3 to 2.5 s and the STI increase from 33% to 40%, at 1000 Hz. Further, the influence of the variation of the audience presence was analyzed, finding out a strict correlation between occupancy percentage and the law of variation of T30. Finally, after the realizations of one of the suggested interventions of acoustic corrections, the Authors conducted a second survey to evaluate the effects of the refurbishment interventions on the acoustic climate. Since the obtained results do not match very well with the design previsions, some hypotheses that can justify such discrepancy are illustrated.

Optimization of frequency lowering algorithms for getting the highest speech intelligibility improvement by hearing loss simulation.

High frequency hearing loss is a growing problem for both children and adults. To overcome this impairment, different frequency lowering methods (FLMs) were tried from 1930s, however no satisfaction was provided up to now. In this study, for getting higher speech intelligibility, eight combinations of FLMs which were designed originally were tried with simulated sounds onto normal hearing subjects. These improvements were calculated by the difference with standard hearing aid method, amplification. High frequency hearing loss was simulated with the combined suprathreshold effects. An offline study was carried out for each subject for determining the significant methods used in modified rhyme test (MRT) (Subjective measure for intelligibility). Significant methods were determined according to their speech intelligibility index (SII) (Objective measure for intelligibility). All different cases were tried under four noisy environments and a noise free environment. Twelve hearing impaired subjects were simulated by hearing loss simulation (HLS). MRT was developed for Turkish language as a first time. As the results of improvements, total 71 cases were statistically significant for twelve subjects. Eighty-three percent success of FLMs was achieved against amplification for being an alternative method of amplification in noisy environments. For four subjects, all significant methods gave higher improvements than amplification. As conclusion, specific method recommendations for different noisy environments were done for each subject for getting more speech intelligibility.

Memory performance on the Auditory Inference Span Test is independent of background noise type for young adults with normal hearing at high speech intelligibility.

Listening in noise is often perceived to be effortful. This is partly because cognitive resources are engaged in separating the target signal from background noise, leaving fewer resources for storage and processing of the content of the message in working memory. The Auditory Inference Span Test (AIST) is designed to assess listening effort by measuring the ability to maintain and process heard information. The aim of this study was to use AIST to investigate the effect of background noise types and signal-to-noise ratio (SNR) on listening effort, as a function of working memory capacity (WMC) and updating ability (UA). The AIST was administered in three types of background noise: steady-state speech-shaped noise, amplitude modulated speech-shaped noise, and unintelligible speech. Three SNRs targeting 90% speech intelligibility or better were used in each of the three noise types, giving nine different conditions. The reading span test assessed WMC, while UA was assessed with the letter memory test. Twenty young adults with normal hearing participated in the study. Results showed that AIST performance was not influenced by noise type at the same intelligibility level, but became worse with worse SNR when background noise was speech-like. Performance on AIST also decreased with increasing memory load level. Correlations between AIST performance and the cognitive measurements suggested that WMC is of more importance for listening when SNRs are worse, while UA is of more importance for listening in easier SNRs. The results indicated that in young adults with normal hearing, the effort involved in listening in noise at high intelligibility levels is independent of the noise type. However, when noise is speech-like and intelligibility decreases, listening effort increases, probably due to extra demands on cognitive resources added by the informational masking created by the speech fragments and vocal sounds in the background noise.

Vowels and intelligibility in dysarthric speech

Diane Kewley-Port’s work in vowel perception under challenging listening conditions and in the relation between vowel perception and production in second language learners has important implications for disordered speech. Vowel space area has been widely used as an index of articulatory working space in speakers with hypokinetic dysarthria related to Parkinson disease (PD), with the assumption that a larger vowel space is associated with higher speech intelligibility. Although many studies have reported acoustic measures of vowels in Parkinson disease, vowel identification and transcription tasks designed to relate changes in production with changes in perception are rarely performed. This study explores the effect of changes in vowel production by six talkers with PD speaking at habitual and loud levels of effort on listener perception. The relation among vowel acoustic measures (including vowel space area and measures of temporal and spectral distinctiveness), vowel identification scores, speech intelligibility ratings, and sentence transcription accuracy for speakers with dysarthria will be discussed.

Measurements of monopolar and bipolar current spreads using forward-masking with a fixed probe

ObjectivesThis research employed a forward-masking paradigm to estimate the current spread of monopolar (MP) and bipolar (BP) maskers, with current amplitudes adjusted to elicit the same loudness. Since the spatial separation between active and return electrodes is smaller in BP than in MP configurations, the BP current spread is more localized and presumably superior in terms of speech intelligibility. Because matching the loudness requires higher current in BP than in MP stimulation, previous forward-masking studies show that BP current spread is not consistently narrower across subjects or electrodes within a subject.MethodsThe present forward-masking measures of current spread differ from those of previous studies by using the same BP probe electrode configuration for both MP and BP masker configurations, and adjusting the current levels of the MP and BP maskers so as to match them in loudness. With this method, the estimate of masker current spread would not be contaminated by differences in probe current spread. Forward masking was studied in four cochlear implant patients, two females and two males, with speech recognition scores higher than 50%; that is, their auditory-nerve survival status was more than adequate to carry out the experiments.ResultsThe data showed that MP and BP masker configurations produce equivalent masking patterns (and current spreads) in three participants. A fourth participant displayed asymmetrical patterns with enhancement rather than masking in some cases, especially when the probe and masker were at the same location.DiscussionThis study showed equivalent masking patterns for MP and BP maskers when the BP masker current amplitude was increased to match the loudness of the MP masker, and the same BP probe configuration is used with both maskers. This finding could help to explain why cochlear implant users often fail to accrue higher speech intelligibility benefit from BP stimulation.

Analysis of the Acoustic Performance of Classrooms in Korea

The basic unit of school is a classroom and the aural environment of the classrooms is essential factor for education purposes. Therefore, many efforts have been undertaken for enhancing the acoustical performance of the classrooms in many countries. As a result, acoustic criteria including reverberation time and background noise level have been established in US and UK for school classrooms depending on the usage and size of the rooms. However, in Korea, there has been little researches concerning the room acoustical investigations of the classrooms. The present study investigates the current situation of the aural environment of the 15 classrooms in Korea including elementary, middle and high schools. The acoustic criteria measured include RT, , STI, SNR and background noise level. As the results, it was found that the background noise levels of the schools adjacent to roads exceed the US and UK standard of 35 dB(A). Also, most schools have so low SNR that they may be interfered by noise, which may affect speech transmission. It was also revealed that some schools have longer RT than the US standard of 0.6 s, but they all have high speech intelligibility.

The effect of bone conduction microphone placement on intensity and spectrum of transmitted speech items

Speech signals can be converted into electrical audio signals using either conventional air conduction (AC) microphone or a contact bone conduction (BC) microphone. The goal of this study was to investigate the effects of the location of a BC microphone on the intensity and frequency spectrum of the recorded speech. Twelve locations, 11 on the talker's head and 1 on the collar bone, were investigated. The speech sounds were three vowels (/u/, /a/, /i/) and two consonants (/m/, /∫/). The sounds were produced by 12 talkers. Each sound was recorded simultaneously with two BC microphones and an AC microphone. Analyzed spectral data showed that the BC recordings made at the forehead of the talker were the most similar to the AC recordings, whereas the collar bone recordings were most different. Comparison of the spectral data with speech intelligibility data collected in another study revealed a strong negative relationship between BC speech intelligibility and the degree of deviation of the BC speech spectrum from the AC spectrum. In addition, the head locations that resulted in the highest speech intelligibility were associated with the lowest output signals among all tested locations. Implications of these findings for BC communication are discussed.

A New Technique for Perceptual Distortion Measure on A Spectro Temporal Auditory Model

Speech processing is used widely in every day’s applications that most people take for granted, such as network wire lines, cellular telephony, telephony system and telephone answering machines. Due to its popularity and increasing of demand, engineers are trying various approaches of improving the process. One of the methods for improving is trying on different methods of filtering techniques. Thus, this instigates an introduction of a filtering technique known as Kalman filtering. In the early days, Kalman filtering was very popular in the research field of navigation because of its magnificent accurate estimation characteristic. Since then, electronic engineers manipulate its advantages to useful purpose in speech processing. Consequently, today it had become a popular filtering technique for estimating and resolving redundant errors containing in speech. A speech pre-processing algorithm is presented to improve the speech intelligibility in noise for the near-end listener. The algorithm improves the intelligibility by optimally redistributing the speech energy over time and frequency for a perceptual distortion measure, which is based on a spectro-temporal auditory model. In contrast o spectral-only models, short-time information is taken into account. As a consequence, the algorithm is more sensitive to transient regions, which will therefore receive more amplification compared to stationary vowels. It is known from literature that changing the vowel-transient energy ratio is beneficial for improving speech intelligibility in noise. Objective intelligibility prediction results show that the proposed method has higher speech intelligibility in noise compared to other reference methods, without modifying the global speech energy.

Comparison of Speech Intelligibility Measures for An Electronic Amplifying Earmuff and An Identical Passive Attenuation Device

The purpose of this study was to identify any differences between speech intelligibility measures obtained with MineEars electronic earmuffs (ProEars, Westcliffe, CO, USA) and the Bilsom model 847 (Sperian Hearing Protection, San Diego, CA, USA), which is a conventional passive-attenuation earmuff. These two devices are closely related, since the MineEars device consisted of a Bilsom 847 earmuff with the addition of electronic amplification circuits. Intelligibility scores were obtained by conducting listening tests with 15 normal-hearing human subject volunteers wearing the earmuffs. The primary research objective was to determine whether speech understanding differs between the passive earmuffs and the electronic earmuffs (with the volume control set at three different positions) in a background of 90 dB(A) continuous noise. As expected, results showed that speech intelligibility increased with higher speech-to-noise ratios; however, the electronic earmuff with the volume control set at full-on performed worse than when it was set to off or the lowest on setting. This finding suggests that the maximum volume control setting for these electronic earmuffs may not provide any benefits in terms of increased speech intelligibility in the background noise condition that was tested. Other volume control settings would need to be evaluated for their ability to produce higher speech intelligibility scores. Additionally, since an extensive electro-acoustic evaluation of the electronic earmuff was not performed as a part of this study, the exact cause of the reduced intelligibility scores at full volume remains unknown.

Cognitive Load During Speech Perception in Noise: The Influence of Age, Hearing Loss, and Cognition on the Pupil Response

The aim of the present study was to evaluate the influence of age, hearing loss, and cognitive ability on the cognitive processing load during listening to speech presented in noise. Cognitive load was assessed by means of pupillometry (i.e., examination of pupil dilation), supplemented with subjective ratings. Two groups of subjects participated: 38 middle-aged participants (mean age = 55 yrs) with normal hearing and 36 middle-aged participants (mean age = 61 yrs) with hearing loss. Using three Speech Reception Threshold (SRT) in stationary noise tests, we estimated the speech-to-noise ratios (SNRs) required for the correct repetition of 50%, 71%, or 84% of the sentences (SRT50%, SRT71%, and SRT84%, respectively). We examined the pupil response during listening: the peak amplitude, the peak latency, the mean dilation, and the pupil response duration. For each condition, participants rated the experienced listening effort and estimated their performance level. Participants also performed the Text Reception Threshold (TRT) test, a test of processing speed, and a word vocabulary test. Data were compared with previously published data from young participants with normal hearing. Hearing loss was related to relatively poor SRTs, and higher speech intelligibility was associated with lower effort and higher performance ratings. For listeners with normal hearing, increasing age was associated with poorer TRTs and slower processing speed but with larger word vocabulary. A multivariate repeated-measures analysis of variance indicated main effects of group and SNR and an interaction effect between these factors on the pupil response. The peak latency was relatively short and the mean dilation was relatively small at low intelligibility levels for the middle-aged groups, whereas the reverse was observed for high intelligibility levels. The decrease in the pupil response as a function of increasing SNR was relatively small for the listeners with hearing loss. Spearman correlation coefficients indicated that the cognitive load was larger in listeners with better TRT performances as reflected by a longer peak latency (normal-hearing participants, SRT50% condition) and a larger peak amplitude and longer response duration (hearing-impaired participants, SRT50% and SRT84% conditions). Also, a larger word vocabulary was related to longer response duration in the SRT84% condition for the participants with normal hearing. The pupil response systematically increased with decreasing speech intelligibility. Ageing and hearing loss were related to less release from effort when increasing the intelligibility of speech in noise. In difficult listening conditions, these factors may induce cognitive overload relatively early or they may be associated with relatively shallow speech processing. More research is needed to elucidate the underlying mechanisms explaining these results. Better TRTs and larger word vocabulary were related to higher mental processing load across speech intelligibility levels. This indicates that utilizing linguistic ability to improve speech perception is associated with increased listening load.

Prediction of optimal conditions for verbal-communication quality in eating establishments

This paper discusses the prediction of verbal-communication quality in eating establishments (EEs). EEs contain talkers and listeners who require high speech intelligibility at their tables, and high speech privacy between tables. Using catt-Acoustic, verbal-communication quality--quantified by speech transmission index (STI)--in models of three existing EEs was predicted. Talker voice-output levels were predicted using an existing empirical model accounting for the Lombard effect. With these, catt-Acoustic predicted impulse responses, speech levels and noise levels at primary and secondary listener positions, and the corresponding STIs. The untreated EEs were first modeled for various talker and listener positions, and occupancies. Then various treated configurations, involving reduced volume, increased absorption and barriers were studied to determine the effectiveness of the treatments. The results suggest that placing barriers around tables can be an effective way to achieve good verbal-communication quality. Increasing the absorption of the room surfaces or decreasing the ceiling height to control reverberation may not be effective. However, increasing the surface absorption and putting barriers around tables may achieve optimal speech conditions in EEs. Subdividing large EEs into smaller ones can also be effective.