Subjective Speech Intelligibility Research Articles

Subjective Speech Intelligibility Drives Noise-Tolerance Domain Use During the Tracking of Noise-Tolerance Test.

Recently, the Noise-Tolerance Domains Test (NTDT) was applied to study the noise-tolerance domains used by young normal-hearing (NH) listeners during noise acceptance decisions. In this study, we examined how subjective speech intelligibility may drive noise acceptance decisions by applying the NTDT on NH and hearing-impaired (HI) listeners at the signal to noise ratios (SNRs) around the Tracking of Noise-Tolerance (TNT) thresholds. A single-blind, within-subjects design with 22 NH and 17 HI older adults was followed. Listeners completed the TNT to determine the average noise acceptance threshold (TNTAve). Then, listeners completed the NTDT at the SNRs of 0, ±3 dB (re: TNTAve) to estimate the weighted noise-tolerance domain ratings (WNTDRs) for each domain criterion. Listeners also completed the Objective and Subjective Intelligibility Difference (OSID) Test to establish the individual intelligibility performance-intensity (P-I) functions of the TNT materials. All test measures were conducted at 75 and 82 dB SPL speech input levels. NH and HI listeners were tested in the unaided mode. The HI listeners were also tested using a study hearing aid. The WNTDRs were plotted against subjective speech intelligibilities extrapolated from individual P-I of the OSID at the SNRs corresponding to NTDT test conditions. Listeners were grouped according to their most heavily weighed domain and a regression analysis was performed against listener demographics as well as TNT and OSID performances to determine which variable(s) affected listener grouping. Three linear mixed effects (LMEs) models were used to examine whether WNTDRs changed with subjective speech intelligibility. All three LMEs found significant fixed effects of domain criteria, subjective intelligibility, and speech input level on WNTDRs. In general, heavier weights were assigned to speech interference and loudness domains at poorer intelligibility levels (<50%) with reversals to distraction and annoyance at higher intelligibility levels (>80%). The comparison between NH and HI-unaided showed that NH listeners assigned greater weights to loudness than the HI-unaided listeners. The comparison between NH and HI-aided groups showed similar weights between groups. The comparison between HI-unaided and HI-aided found that HI listeners assigned lower weights to speech interference and greater weights to loudness when tested in aided compared with unaided modes. In all comparisons, loudness was weighed heavier at the 82 dB SPL input level than at the 75 dB SPL input level with greater weights to annoyance in the NH versus HI-unaided comparison and lower weights to distraction in the HI-aided versus HI-unaided comparison. A generalized linear model determined that listener grouping was best accounted for by subjective speech intelligibility estimated at TNTAve. The domain criteria used by listeners were driven by their subjective speech intelligibility regardless of their hearing status (i.e., NH versus HI). In general, when subjective intelligibility was poor, the domains of speech interference and loudness were weighed the heaviest. As subjective intelligibility improved, the weightings on annoyance and distraction increased. Furthermore, a listener's criterion for >90% subjective speech understanding at the TNTAve may allow one to profile the listener.

Measuring Objective and Subjective Intelligibility Using Speech Materials from the Tracking of Noise Tolerance (TNT) Test.

The results of tests measuring objective speech intelligibility are similar to those measuring subjective speech intelligibility using speech materials with minimal context. It is unclear if such is the case with contextual materials. This article compares objective and subjective intelligibility difference (OSID) between normal hearing (NH) and hearing impaired (HI) listeners in the unaided and aided modes using speech materials adapted from the Tracking of Noise Tolerance (TNT) test. Single-blind within-subjects design. Twenty-four NH and 17 HI older adults. Listeners completed the objective and subjective intelligibility measures at 75 and 82 dB sound pressure level (SPL) speech input levels. Five signal-to-noise ratios were tested to generate the objective and subjective speech intelligibility performance intensity (P-I) functions. Both NH and HI listeners were tested in the unaided mode. The HI listeners were also tested using their own hearing aids (HAs). Objective and subjective speech reception thresholds at a 50% criterion (SRT50s) were estimated from each individual P-I function. The difference between the objective SRT50 and subjective SRT50 was used to estimate OSID. Objective and subjective SRT50s were significantly better in NH than in HI listeners (chi-square(1) = 26.29, p < 0.001) at each speech input level in the unaided mode. However, there was a significant interaction between listener group and intelligibility type (chi-square(1) = 9.43, p = 0.002) where SRT50s were lower for subjective than objective P-I functions only in the HI group. The SRT50s of HI listeners were also affected by hearing mode, where both objective and subjective intelligibility was improved when HI listeners were tested while wearing their own HAs. In general, objective and subjective SRT50s showed moderate-to-strong correlations across most combinations of listener groups and test conditions (r = 0.59-0.86, p < 0.01) except for HI listeners tested with their own HAs (r = 0.39, p = 0.128). Similar objective and subjective intelligibility was observed in NH listeners but better subjective intelligibility than objective intelligibility was noted in HI listeners when tested in the unaided and aided modes.

Performance of older normal-hearing listeners on the tracking of noise tolerance (TNT) test

Objective To gather preliminary reference data on older normal-hearing (NH) adults for the refined Tracking of Noise Tolerance (TNT) test. Design Within-subject repeated measures. Participants were tested on the TNT in the sound-field and under headphones. In the sound-field, speech stimuli were presented at 75 dB SPL and 82 dB SPL from 0° with a speech-shaped noise presented either from 0° or 180° at a level controlled by the participants. The order of signal level, mode of presentation, noise azimuth, and TNT passages were counterbalanced across listeners. Testing was repeated for one condition after 1–3 weeks to estimate within-session and between-session reliability. Study sample Twenty-five NH listeners (51–82 yrs of age). Results Mean TNT scores (TNTAve) were about 4 dB at a speech input of 75 dB SPL and 3 dB at 82 dB SPL. The TNTAve was similar between the headphone and sound-field presentations in the co-located noise. TNTAve scores measured with noise-back were about 1 dB better than those measured from the front. The 95% confidence intervals of absolute test-retest differences were about 1.2 dB within-session and 2.0 dB between sessions. Conclusions The refined TNT may be a reliable tool to measure noise acceptance and subjective speech intelligibility.

The effects of estimation accuracy, estimation approach, and number of selected channels using formant-priority channel selection for an "n-of-m" sound processing strategy for cochlear implants.

Previously, selection of l channels was prioritized according to formant frequency locations in an l-of-n-of-m-based signal processing strategy to provide important voicing information independent of listening environments for cochlear implant (CI) users. In this study, ideal, or ground truth, formants were incorporated into the selection stage to determine the effect of accuracy on (1) subjective speech intelligibility, (2) objective channel selection patterns, and (3) objective stimulation patterns (current). An average +11% improvement (p < 0.05) was observed across six CI users in quiet, but not for noise or reverberation conditions. Analogous increases in channel selection and current for the upper range of F1 and a decrease across mid-frequencies with higher corresponding current, were both observed at the expense of noise-dominant channels. Objective channel selection patterns were analyzed a second time to determine the effects of estimation approach and number of selected channels (n). A significant effect of estimation approach was only observed in the noise and reverberation condition with minor differences in channel selection and significantly decreased stimulated current. Results suggest that estimation method, accuracy, and number of channels in the proposed strategy using ideal formants may improve intelligibility when corresponding stimulated current of formant channels are not masked by noise-dominant channels.

On the deficiency of intelligibility metrics as proxies for subjective intelligibility

A recent trend in deep neural network (DNN)-based speech enhancement consists of using intelligibility and quality metrics as loss functions for model training with the aim of achieving high subjective speech intelligibility and perceptual quality in real-life conditions. In this study, we analyze a variety of loss functions, including some based on state-of-the-art intelligibility and quality metrics, to train an end-to-end speech enhancement system based on a fully convolutional neural network. The loss functions include perceptual metric for speech quality evaluation (PMSQE), scale-invariant signal-to-distortion ratio (SI-SDR), SI-SDR integrating speech pre-emphasis, short-time objective intelligibility (STOI), extended STOI (ESTOI), spectro-temporal glimpsing index (STGI), and a composite loss function combining STGI and SI-SDR. While DNNs trained with these loss functions produce notable speech intelligibility (and quality) gains according to pertinent objective metrics, we conduct a subjective intelligibility test that contradicts this result, showing no intelligibility improvement. From the results of this study, our conclusion is twofold: (1) subjective intelligibility evaluation is currently not replaceable by objective intelligibility evaluation, and (2) both the development of meaningful intelligibility metrics and DNN-based speech enhancement systems that can consistently improve the intelligibility of noisy speech for human listening remain open problems.

The Attenuation and Speech Intelligibility of the Gecko Helmet Used by the Royal National Lifeboat Institution

The RNLI operates a fleet of B Class Inshore Lifeboats around the UK and Ireland. To help keep their volunteers safe, crew members wear the Gecko helmet as part of their issued personal protective equipment. Anecdotal evidence from lifeboat crew has shown that wearing the helmet interfered with the ability of volunteers to follow normal conversations and commands given under non-amplified situations. Objective measurements were undertaken at varying angles of incidence to quantify the attenuation of sound provided by the helmet at the location of the human ear along with objective and subjective measurements of the intelligibility of speech while wearing the Gecko helmet. Analysis of data showed that the Gecko helmet provides limited noise attenuation at frequencies above 1kHz. Results of objective and subjective speech intelligibility tests show a reduction in the speech transmission index score when the helmet was introduced into the experiment.

Spatial release from masking in varying spatial acoustic under higher order ambisonic-based sound reproduction system

A previous study found that spatial release from masking (SRM) could be observed under virtual reverberant environments using a first order Ambisonic-based sound reproduction system, however, poor localisation accuracy made it difficult to examine effect of varying reverberation time on SRM. The present study follows on using higher order Ambisonics (HOA) to examine how benefits from SRM vary in different spatial acoustics. Subjective speech intelligibility was measured where four room acoustics:reverberation time (RT)= 0.7 s (clarity (C50)= 16 dB, 7 dB); RT= 1.8 s (C50= 8 dB, 2 dB) were simulated via a third order Ambisonic system with a 16 channel spherical loudspeaker array. The masker was played from 8 azimuthal angles (0, +-45, +-90, +-135, 180 degrees) while the target speech was played from 0 degree. The listeners are deemed to benefit from SRM if their intelligibility scores were higher when the masker comes from a different angle than that of the target. We found while listeners could benefit from SRM at C50 = 16 dB and 8 dB, the benefit starts to diminish at C50 = 7 dB, and listeners could no longer benefit from SRM at C50 = 2 dB.

Effects of Finding the Speech-Language Pathologist Likeable on Postlaryngectomy Speech Intelligibility Outcomes

Introduction: Speech-language pathologists (SLPs) work with patients after total laryngectomy (TL) to regain verbal communication. The influence of the quality of the therapeutic relationship on the success of TL voice rehabilitation in terms of speech intelligibility is not known. Finding each other likeable is an important factor in establishing and maintaining interpersonal relationships in everyday life. The fit of therapist and client is relevant to the therapeutic relationship. The purpose of this study therefore was to assess the association between the degree of SLPs’ likeability ratings and postlaryngectomy speech intelligibility. Methods: In a multicentre prospective cohort study, participants rated their SLPs’ likeability after finishing TL rehabilitation. Speech intelligibility was measured objectively with the Post-Laryngectomy Telephone Intelligibility Test and subjectively with the Questionnaire for Adjustment after Laryngectomy. The association of SLPs’ likeability with speech intelligibility was analysed using hierarchical logistic regression, expressed with odds ratios (OR) with corresponding 95% confidence intervals (CI). Results: Altogether 124 patients from 13 institutions participated. The degree of finding the SLP likeable was not significantly associated with objective speech intelligibility (OR 1.30; 95% CI 0.78–2.18; p = 0.32) or subjective speech intelligibility (OR 1.01; 95% CI 0.60–1.72; p = 0.96) after controlling for age, sex and education factors. Discussion/Conclusion: In this patient cohort, there was no evidence for an association between ratings of SLPs’ likeability and speech intelligibility outcomes after rehabilitation. Future studies could consider the use of alternative instruments for measuring likeability.

Transient Noise Reduction Using a Deep Recurrent Neural Network: Effects on Subjective Speech Intelligibility and Listening Comfort.

A deep recurrent neural network (RNN) for reducing transient sounds was developed and its effects on subjective speech intelligibility and listening comfort were investigated. The RNN was trained using sentences spoken with different accents and corrupted by transient sounds, using the clean speech as the target. It was tested using sentences spoken by unseen talkers and corrupted by unseen transient sounds. A paired-comparison procedure was used to compare all possible combinations of three conditions for subjective speech intelligibility and listening comfort for two relative levels of the transients. The conditions were: no processing (NP); processing using the RNN; and processing using a multi-channel transient reduction method (MCTR). Ten participants with normal hearing and ten with mild-to-moderate hearing loss participated. For the latter, frequency-dependent linear amplification was applied to all stimuli to compensate for individual audibility losses. For the normal-hearing participants, processing using the RNN was significantly preferred over that for NP for subjective intelligibility and comfort, processing using the RNN was significantly preferred over that for MCTR for subjective intelligibility, and processing using the MCTR was significantly preferred over that for NP for comfort for the higher transient level only. For the hearing-impaired participants, processing using the RNN was significantly preferred over that for NP for both subjective intelligibility and comfort, processing using the RNN was significantly preferred over that for MCTR for comfort, and processing using the MCTR was significantly preferred over that for NP for comfort.

Combination and Comparison of Sound Coding Strategies Using Cochlear Implant Simulation With Mandarin Speech.

Three cochlear implant (CI) sound coding strategies were combined in the same signal processing path and compared for speech intelligibility with vocoded Mandarin sentences. The three CI coding strategies, biologically-inspired hearing aid algorithm (BioAid), envelope enhancement (EE), and fundamental frequency modulation (F0mod), were combined with the advanced combination encoder (ACE) strategy. Hence, four singular coding strategies and four combinational coding strategies were derived. Mandarin sentences with speech-shape noise were processed using these coding strategies. Speech understanding of vocoded Mandarin sentences was evaluated using short-time objective intelligibility (STOI) and subjective sentence recognition tests with normal-hearing listeners. For signal-to-noise ratios at 5 dB or above, the EE strategy had slightly higher average scores in both STOI and listening tests compared to ACE. The addition of EE to BioAid slightly increased the mean scores for BioAid+EE, which was the combination strategy with the highest scores in both objective and subjective speech intelligibility. The benefits of BioAid, F0mod, and the four combinational coding strategies were not observed in CI simulation. The findings of this study may be useful for the future design of coding strategies and related studies with Mandarin.

Is the cochlear implant a successful long-term solution for single-sided deaf and asymmetric hearing-impaired patients?

PurposeWe investigated the long-term results of cochlear implant (CI) recipients with asymmetric hearing loss (AHL) or single-sided deafness (SSD). We focused on wearing behavior, audiometric hearing rehabilitation, and subjective benefits of the CI. CI is expected to improve audiological results, subjective hearing perception, and tinnitus burden.MethodsSpeech recognition in background noise and sound localization were assessed preoperatively and after at least six years of CI experience. Validated questionnaires determined the subjective benefit of CI use and the subjective evaluation of tinnitus.ResultsOver 80% of the included AHL and SSD CI recipients used their CI between 6 and 10 h daily; four subjects with SSD were non-users. Speech recognition in background noise and sound localization improved significantly compared with the unaided preoperative situation. Additionally, CI improved subjective speech intelligibility and spatial hearing impression while reducing tinnitus burden.ConclusionSubjects with AHL and SSD benefit from CI, subjectively and audiologically. Cochlear implant is a successful long-term treatment for AHL and SSD.

Effects of acoustical descriptors on speech intelligibility in Hong Kong classrooms

This study investigated the effects of classroom acoustics on speech intelligibility in secondary school and university classrooms. Speech intelligibility tests were conducted in 9 secondary school classrooms and 18 university classrooms and the acoustical measurements were performed in these classrooms. Subjective speech intelligibility tests were obtained from phonetically balanced (PB) word lists on a total of 672 students and acoustic descriptors such as signal-to-noise ratio (SNR), early decay time (EDT), and sound clarity (C80) were conducted in different listening positions in each classroom. The relationships between speech intelligibility scores (SI) and acoustical descriptors were fitted based on non-linear curve fitting regression models. The “S” form regression model was selected with modification as the basic regression equation to describe the effects of SNR on speech intelligibility. The combination effects of SNR with reverberation condition and sound clarity condition on speech intelligibility were investigated. The impact of different age groups and linguistic environment on speech intelligibility were discussed.The results reveal that SI increases with the increase of SNR value for all age groups. The results indicate that nearly 0.06 s increasing in EDT values will result in a 1% decrease in SI. Furthermore, the results also suggest that a 1 dB increasing in C80 values will result in a 1.23% increase in speech intelligibility scores. The SI increases as the age increases under the same SNR condition. The speech intelligibility scores are always lower than the comparison research results with a constant reverberation value as well as sound clarity value for an equal SNR value.

Improving Patient Voice Intelligibility by using a Euclidian Distance-Based Approach to Improve Voice Assistant Accuracy

Voice assistance (VA) is gaining domestic consumer attention in a variety of products, such as Amazon Alexa, Google Home, Apple’s Siri, and Microsoft’s Cortana. Furthermore, VA has recently shown its usefulness and ability to improve inpatient experience in hospitals and clinics. Nevertheless, none of the VA products has an accuracy rate greater than 90%. The accuracy decreases even more in noisy or public environments. Hence, improving VA accuracy in noisy environments requires a speech signal algorithm with good quality and intelligibility. There is great interest in developing an objective intelligibility measure that shows maximum correlation with subjective speech intelligibility and that can measure the effect of speech enhancement algorithms on the processing of noisy speech signals. In this paper, Euclidian distance-based speech intelligibility prediction is proposed to measure the correlation with subjective intelligibility in different noisy environments. This paper also presents a comparative analysis and general background research in speech intelligibility improvement. The results show that no single algorithm is effective in improving the intelligibility of speech signals.

The speech intelligibility and applicability of the speech transmission index in large spaces

This paper aims to explore the influence factors of speech intelligibility and the applicability of the speech transmission index (STI) in large spaces, where the sound energy is unevenly distributed and non-exponentially decays. The subjective speech intelligibility tests were conducted in Mandarin (China mainland) in two large spaces with volumes of 97,000 m3 and 246,000 m3. Objective indicators such as the Reverberation Time (RT), Early Decay Time (EDT), Definition (D50), and Speech Transmission Index (STI) under different signal-to-noise ratio (SNRs) were also measured in these two spaces. The results showed that both the SNR and room acoustics had significant effect on the speech intelligibility in these two spaces, but the effect of room acoustics on speech intelligibility was also affected by SNR. The speech intelligibility scores significantly increased with the increase in SNR when the SNR was less than 14.4 dB. In terms of room acoustics, D50 was more relevant to speech intelligibility than RT and EDT in these two large spaces when SNR ranged from −5dB to 15 dB. The STI value in large spaces should not be used as in ordinary spaces to evaluate the speech intelligibility. Based on the tests in this paper, the corresponding relation between STI and speech intelligibility in large spaces was modified, and a new rating threshold of STI was also proposed according to the revised relation, which indicated a necessity to modify the rating criteria of using STI to predict speech intelligibility in large spaces.

Comparative study of acoustical indices and speech perception of students in two primary school classrooms with an acoustical treatment

Objective acoustical indices and subjective speech perception of students were investigated and compared in two different classrooms before and after an acoustical treatment. The treatment in the present study refers to the installation of sound absorption materials on the ceilings in two primary school classrooms. The results showed that the reverberation time after acoustical treatment was reduced, while the early-to-late sound ratio and the speech transmission index were improved. The subjective speech intelligibility in classrooms was also enhanced. Subjective survey showed that most of the students preferred learning in the classrooms after acoustical treatment. They reported that higher concentration is because the speeches of the teachers were much clear in the treated classroom. However, subjective speech perception of the students has some discrepancies between the two classrooms. The perception of the students in Classroom A was better than that in Classroom B. Sixty-two percent (61.9%) of the students prefer to study in the classroom A after acoustical treatment, but only 46.3% of the students prefer to study in the Classroom B. Results show that exceeding low-frequency reverberation in the classroom may significantly mask speech signals and have effect on the student’s speech perception in classroom, even though the traditional criteria of RT and C50 at mid-frequencies are similar.

Objective Intelligibility Assessment by Automated Segmental and Suprasegmental Listening Error Analysis.

Purpose Subjective speech intelligibility assessment is often preferred over more objective approaches that rely on transcript scoring. This is, in part, because of the intensive manual labor associated with extracting objective metrics from transcribed speech. In this study, we propose an automated approach for scoring transcripts that provides a holistic and objective representation of intelligibility degradation stemming from both segmental and suprasegmental contributions, and that corresponds with human perception. Method Phrases produced by 73 speakers with dysarthria were orthographically transcribed by 819 listeners via Mechanical Turk, resulting in 63,840 phrase transcriptions. A protocol was developed to filter the transcripts, which were then automatically analyzed using novel algorithms developed for measuring phoneme and lexical segmentation errors. The results were compared with manual labels on a randomly selected sample set of 40 transcribed phrases to assess validity. A linear regression analysis was conducted to examine how well the automated metrics predict a perceptual rating of severity and word accuracy. Results On the sample set, the automated metrics achieved 0.90 correlation coefficients with manual labels on measuring phoneme errors, and 100% accuracy on identifying and coding lexical segmentation errors. Linear regression models found that the estimated metrics could predict a significant portion of the variance in perceptual severity and word accuracy. Conclusions The results show the promising development of an objective speech intelligibility assessment that identifies intelligibility degradation on multiple levels of analysis.

Predicting individual speech intelligibility from the cortical tracking of acoustic- and phonetic-level speech representations

ObjectiveTo objectively measure speech intelligibility of individual subjects from the EEG, based on cortical tracking of different representations of speech: low-level acoustical, higher-level discrete, or a combination. To compare each model's prediction of the speech reception threshold (SRT) for each individual with the behaviorally measured SRT. MethodsNineteen participants listened to Flemish Matrix sentences presented at different signal-to-noise ratios (SNRs), corresponding to different levels of speech understanding. For different EEG frequency bands (delta, theta, alpha, beta or low-gamma), a model was built to predict the EEG signal from various speech representations: envelope, spectrogram, phonemes, phonetic features or a combination of phonetic Features and Spectrogram (FS). The same model was used for all subjects. The model predictions were then compared to the actual EEG of each subject for the different SNRs, and the prediction accuracy in function of SNR was used to predict the SRT. ResultsThe model based on the FS speech representation and the theta EEG band yielded the best SRT predictions, with a difference between the behavioral and objective SRT below 1 decibel for 53% and below 2 decibels for 89% of the subjects. ConclusionA model including low- and higher-level speech features allows to predict the speech reception threshold from the EEG of people listening to natural speech. It has potential applications in diagnostics of the auditory system.

Tracking of Noise Tolerance to Predict Hearing Aid Satisfaction in Loud Noisy Environments.

A method that tracked tolerable noise level (TNL) over time while maintaining subjective speech intelligibility was reported previously. Although this method was reliable and efficacious as a research tool, its clinical efficacy and predictive ability of real-life hearing aid satisfaction were not measured. The study evaluated an adaptive method to estimate TNL using slope and variance of tracked noise level as criteria in a clinical setting. The relationship between TNL and subjective hearing aid satisfaction in noisy environments was also investigated. A single-blinded, repeated-measures design. Seventeen experienced hearing aid wearers with bilateral mild-to-moderately-severe sensorineural hearing loss. Participants listened to 82-dB SPL continuous speech and tracked the background noise level that they could "put up with" while subjectively understanding >90% of the speech material. Two trials with each babble noise and continuous speech-shaped noise were measured in a single session. All four trials were completed aided using the participants' own hearing aids. The stimuli were presented in the sound field with speech from 0° and noise from the 180° azimuth. The instantaneous tolerable noise level was measured using a custom program and scored in two ways; the averaged TNL (aTNL) over the 2-min trial and the estimated TNL (eTNL) as soon as the listeners reached a stable noise estimate. Correlation between TNL and proportion of satisfied noisy environments was examined using the MarkeTrak questionnaire. All listeners completed the tracking of noise tolerance procedure within 2 min with good reliability. Sixty-five percent of the listeners yielded a stable noise estimate after 59.9 sec of actual test time. The eTNL for all trials was 78.6 dB SPL (standard deviation [SD] = 4.4 dB). The aTNL for all trials was 78.0 dB SPL (SD = 3.3 dB) after 120 sec. The aTNL was 79.2 dB SPL (SD = 5.4 dB) for babble noise and 77.0 dB SPL (SD = 5.9 dB) for speech-shaped noise. High within-session test-retest reliability was evident. The 95% confidence interval was 1.5 dB for babble noise and 2.8 dB for continuous speech-shaped noise. No significant correlation was measured between overall hearing aid satisfaction and the aTNL (ρ = 0.20 for both noises); however, a significant relationship between aTNL and proportion of satisfied noisy situations was evident (ρ = 0.48 for babble noise and ρ = 0.55 for speech-shaped noise). The eTNL scoring method yielded similar results as the aTNL method although requiring only half the time for 65% of the listeners. This time efficiency, along with its reliability and the potential relationship between TNL and hearing aid satisfaction in noisy listening situations suggests that this procedure may be a good clinical tool to evaluate whether specific features on a hearing aid would improve noise tolerance and predict wearer satisfaction with the selected hearing aid in real-life loud noisy situations. A larger sample of hearing aid wearers is needed to further validate these potential uses.

Subjective Self-Rated Speech Intelligibility and Quality of Life in Patients with Parkinson’s Disease in a Malaysian Sample

Objectives: This study compared the Quality of Life (QOL) of individuals with Parkinson’s disease (PD) with a healthy control from a sample in Malaysia. The relationship between the QOL, age, gender, duration of disease, self-rated speech scale, and life satisfaction level was examined. Methods: Fifty-four individuals with PD and 55 non-PD control adults completed the Parkinson’s Disease Questionnaire-39 (PDQ-39) and rated their speech intelligibility [Self-Rated Speech Scale (SRSS)] and life satisfaction level [Life Satisfaction Scale (LSS)] on a 10-point scale. Results: PD participants reported significantly lower QOL scores than the control group (Mann-Whitney U). Although there was no significant relationship between the QOL, age, gender, and duration of disease for PD participants, moderately significant correlations were observed between QOL, LSS and SRSS ratings (Spearman correlation). Individuals with PD showed lower QOL and higher concern in their communication than the control group. Conclusion: Identifying patients’ perception of their speech performance could help clinicians to better understand patients’ needs when delivering speech therapy services.

Exploiting alternative acoustic sensors for improved noise robustness in speech communication

This study investigates the use of non-conventional body-conductive acoustic sensors in human-human speech communication and automatic speech recognition. The body-conductive sensors are directly attached to the speaker and receive the uttered speech through the skin and bones, resulting in higher robustness against environmental noise. In this study, a throat microphone, an ear bone microphone, and a standard microphone were evaluated using subjective speech intelligibility tests and automatic speech recognition experiments. In addition to the use of these sensors on their own, several methods were also applied for sensor integration, thereby achieving higher recognition rates. Namely, multi-stream hidden Markov model (HMM) decision fusion, and late fusion methods were used to integrate several sensors. By using late fusion, a 40% relative recognition rate improvement in a noisy environment, and a 24% relative recognition rate improvement in a clean environment were achieved. In the case of late fusion, a novel adaptive weighting method was introduced that does not require any pre-adjustment of the weights. In this study, a technique to automatically segment noisy speech data by using a body-conductive sensor in conjunction with the desired microphone during recording is presented. The Lombard effect phenomenon when using body-conductive acoustic sensors was also investigated.