Presence Of Speech-shaped Noise Research Articles

Neural Envelope Processing at Low Frequencies Predicts Speech Understanding of Children With Hearing Loss in Noise and Reverberation.

Children with hearing loss experience greater difficulty understanding speech in the presence of noise and reverberation relative to their normal hearing peers despite provision of appropriate amplification. The fidelity of fundamental frequency of voice (f0) encoding-a salient temporal cue for understanding speech in noise-could play a significant role in explaining the variance in abilities among children. However, the nature of deficits in f0 encoding and its relationship with speech understanding are poorly understood. To this end, we evaluated the influence of frequency-specific f0 encoding on speech perception abilities of children with and without hearing loss in the presence of noise and/or reverberation. In 14 school-aged children with sensorineural hearing loss fitted with hearing aids and 29 normal hearing peers, envelope following responses (EFRs) were elicited by the vowel /i/, modified to estimate f0 encoding in low (<1.1 kHz) and higher frequencies simultaneously. EFRs to /i/ were elicited in quiet, in the presence of speech-shaped noise at +5 dB signal to noise ratio, with simulated reverberation time of 0.62 sec, as well as both noise and reverberation. EFRs were recorded using single-channel electroencephalogram between the vertex and the nape while children watched a silent movie with captions. Speech discrimination accuracy was measured using the University of Western Ontario Distinctive Features Differences test in each of the four acoustic conditions. Stimuli for EFR recordings and speech discrimination were presented monaurally. Both groups of children demonstrated a frequency-dependent dichotomy in the disruption of f0 encoding, as reflected in EFR amplitude and phase coherence. Greater disruption (i.e., lower EFR amplitudes and phase coherence) was evident in EFRs elicited by low frequencies due to noise and greater disruption was evident in EFRs elicited by higher frequencies due to reverberation. Relative to normal hearing peers, children with hearing loss demonstrated: (a) greater disruption of f0 encoding at low frequencies, particularly in the presence of reverberation, and (b) a positive relationship between f0 encoding at low frequencies and speech discrimination in the hardest listening condition (i.e., when both noise and reverberation were present). Together, these results provide new evidence for the persistence of suprathreshold temporal processing deficits related to f0 encoding in children despite the provision of appropriate amplification to compensate for hearing loss. These objectively measurable deficits may underlie the greater difficulty experienced by children with hearing loss.

Intelligibility of Noise-Adapted and Clear Speech in Energetic and Informational Maskers for Native and Nonnative Listeners.

This study explored clear speech (CS) and noise-adapted speech (NAS) intelligibility benefits for native and nonnative English listeners. It also examined how the two speaking style adaptations interact with maskers that vary from purely energetic to largely informational at different signal-to-noise ratios (SNRs). Materials consisted of 40 sentences produced by 10 young adult talkers in a conversational and a clear speaking style under two conditions: (a) in quiet and (b) in response to speech-shaped noise (SSN) played over headphones (NAS). Young adult native (Experiment 1) and nonnative (Experiment 2) English listeners heard target sentences presented in two-talker (2T) babble, six-talker (6T) babble, or SSN and at an "easier" and a "harder" SNR. When talkers produced CS and NAS, word recognition accuracy was significantly improved for both listener groups. The largest intelligibility benefit was obtained for the CS produced in response to noise (CS+NAS). Overall accuracy was highest in 2T babble. Accuracy was higher in SSN than in 6T babble for nonnative listeners at both levels of listening difficulty but only at a more difficult SNR for native listeners. Listeners benefited from CS and NAS most in the presence of SSN and least in 2T babble. When SNRs were the same for the two listener groups, native listeners outperformed nonnative listeners in almost all listening conditions, but nonnative listeners benefited more from CS and NAS in 6T babble than native listeners did. Combined speaking style enhancements, CS+NAS, provided the largest intelligibility increases for native and nonnative listeners in all listening conditions. The results add to the body of evidence supporting speech-oriented, behavioral therapy techniques for maximizing speech intelligibility in everyday listening situations.

Efficacy of Computer-Based Noise Desensitization Training in Children With Speech-in-Noise Deficits.

Purpose Auditory training is known to improve the speech-perception-in-noise (SPIN) skills in children with auditory processing disorders (APDs); however, conventional non-computer-based trainings are dreary and nonmotivating. This study intended to develop a computer-based noise desensitization training module in Indian English and test the efficacy of the same in children with APD having SPIN deficits. There are no such modules available at present to be used in Indian children. Method The module incorporated words-in-noise training, with monosyllables and trisyllables as target words in the presence of speech-shaped noise as well as multitalker babble at different signal-to-noise ratios ranging from +20 to -4 dB SNR. The study included 20 children with SPIN deficits diagnosed with APD and who were recruited randomly to the experimental group who received the training (n = 10; M age = 11.1 years) and to the control group who did not receive any training (n = 10; M age = 11.4 years). The outcome measures included behavioral measures and electrophysiological measures (auditory long latency responses for speech in quiet and noise). Results Compared to children in the control group, children in the experimental group exhibited improvements in the SPIN scores as well as temporal processing measures after training. Dichotic consonant-vowel tests and auditory memory and sequencing tests did not yield significant improvement post training. Also, auditory long latency responses in quiet and noise revealed significant reduction in the amplitudes after training. No significant difference was noted in the latencies of auditory long latency responses post training. Control group participants did not show any significant difference in any of the measures between the initial and follow-up evaluations. Conclusions The developed training module was successful in fetching improvements in behavioral and electrophysiological measures. It would be a great addition to the evidence-based rehabilitation support inventory for children with SPIN deficits in India. Supplemental Material https://doi.org/10.23641/asha.14551041.

Effect of masks on speech intelligibility in auralized classrooms.

This study explored the effects of wearing face masks on classroom communication. The effects of three different types of face masks (fabric, surgical, and N95 masks) on speech intelligibility (SI) presented to college students in auralized classrooms were evaluated. To simulate realistic classroom conditions, speech stimuli were presented in the presence of speech-shaped noise with a signal-to-noise ratio of +3 dB under two different reverberation times (0.4 s and 3.1 s). The use of fabric masks yielded a significantly greater reduction in SI compared to the other masks. Therefore, surgical masks or N95 masks are recommended in teaching environments.

Speech intelligibility in auralized classrooms when the talker is wearing a face mask

This study explored the effects of wearing face masks on classroom communication. We evaluated the effects of three different types of face masks (fabric, surgical and N95 masks) on speech intelligibility presented to college students in auralized classrooms. To simulate realistic classroom conditions, speech stimuli were presented in the presence of speech-shaped noise with a signal-to-noise ratio of + 3 dB under two different reverberation times (0.4 s and 3.1 s). The use fabric masks yielded significantly greater reduction in speech intelligibility compared to the other masks. Therefore, surgical masks or N95 masks are strongly recommended in teaching environments.

Talker depressive symptoms affect intelligibility of noise-adapted speech

A critical aspect of communicative competence is talker's ability to adapt the clarity of their speech in response to the challenges of a particular communicative situation. Previous work showed that talkers with elevated depressive symptoms were less successful in producing listener-oriented intelligibility-enhancing clear speech. The present study investigated intelligibility of speech produced in response to environmental noise (noise-adapted speech, NAS) in talkers with high depressive (HD) and low depressive (LD) symptoms. Sentence intelligibility was examined in the presence of speech-shaped noise (SSN) and 1-talker (1T) competing speech. Results revealed that NAS increased intelligibility for both talker groups and in both maskers. NAS intelligibility benefit was smaller for HD talkers than LD talkers in the SSN condition. Acoustic analyses showed that NAS was characterized by a decreased speaking rate, increased F0 mean and range, and increased energy in the 1–3 kHz range. Talkers with HD symptoms however produced these modifications significantly less compared to talkers with LD symptoms. Results provide further evidence that elevated depressive symptoms impact speaking style adaptations in response to various communicative barriers leading to lower intelligibility. The results have the potential to aid in clinical decision making for individuals with depression.

Background Speech Disrupts Working Memory Span in 5-Year-Old Children

The present study tested the effects of background speech and nonspeech noise on 5-year-old children's working memory span. Five-year-old typically developing children (range = 58.6 to 67.6 months; n = 94) completed a modified version of the Missing Scan Task, a missing-item working memory task, in quiet and in the presence of two types of background noise: male two-talker speech and speech-shaped noise. The two types of background noise had similar spectral composition and overall intensity characteristics but differed in whether they contained verbal content. In Experiments 1 and 2, children's memory span (i.e., the largest set size of items children successfully recalled) was subjected to analyses of variance designed to look for an effect of listening condition (within-subjects factor: quiet, background noise) and an effect of background noise type (between-subjects factor: two-talker speech, speech-shaped noise). In Experiment 1, children's memory span declined in the presence of two-talker speech but not in the presence of speech-shaped noise. This result was replicated in Experiment 2 after accounting for a potential effect of proactive interference due to repeated administration of the Missing Scan Task. Background speech, but not speech-shaped noise, disrupted working memory span in 5-year-old children. These results support the idea that background speech engages domain-general cognitive processes used during the recall of known objects in a way that speech-shaped noise does not.

Effect of speech-shaped noise on the recognition of malayalam and english consonants by malayalam listeners.

Objective:The present study compared recognition of native and non-native consonants in quiet and noise among native speakers of Malayalam.Methods and Material:Fifteen native speakers of Malayalam who had English as the medium of instruction at school participated in the study. Stimuli comprised of 16 vowel-consonants-vowel nonsense syllables spoken by eight native speakers of Malayalam (native consonants) and eight native speakers of American English (non-native consonants). Recognition of native and non-native consonants was studied in quiet and in the presence of speech-shaped noise at signal-to-noise ratios (SNRs) of 8 dB, 0 dB, and −8 dB. The consonant recognition task was carried out as 16-alternative forced-choice procedure, and the responses were stored as confusion matrix.Results:In favourable listening condition (i.e., quiet and 8 dB SNR), the recognition score for native consonants was greater than non-native consonants. In contrast, at 0 dB SNR and −8 dB SNR, the recognition score of non-native consonants was greater than native consonants. Information transfer analysis revealed that the transfer of information was highest for consonant feature manner of articulation and lowest for voicing, across listening conditions for both native and non-native consonants.Conclusions:Recognition of native and non-native consonants were affected differently in the presence of speech-shaped noise among native speakers of Malayalam. In favourable listening condition, recognition of native consonants was better than non-native consonants. However, in challenging listening condition, non-native consonants were found to be recognised better than native consonants.

Comparing human and automatic speech recognition in simple and complex acoustic scenes

Former comparisons of human speech recognition (HSR) and automatic speech recognition (ASR) have shown that humans outperform ASR systems in nearly all speech recognition tasks. However, recent progress in ASR has led to substantial improvements of recognition accuracy, and it is therefore unclear how large the task-dependent human-machine gap still remains. This paper investigates this gap between HSR and ASR based on deep neural networks (DNNs) in different acoustic conditions, with the aim of comparing differences and identifying processing strategies that should be considered in ASR. We find that DNN-based ASR reaches human performance for single-channel, small-vocabulary tasks in the presence of speech-shaped noise and in multi-talker babble noise, which is an important difference to previous human-machine comparisons: The speech reception threshold, i.e., the signal-to-noise ratio with 50% word recognition rate is at about −7 to −8 dB both for HSR and ASR. However, in more complex spatial scenes with diffuse noise and moving talkers, the SRT gap amounts to approximately 12 dB. Based on cross comparisons that use oracle knowledge (e.g., the speakers’ true position), incorrect responses are attributed to localization errors or missing pitch information to distinguish between speakers with different gender. In terms of the SRT, localization errors and missing spectral information amount to 2.1 and 3.2 dB, respectively. The comparison hence identifies specific components in ASR that can profit from learning from auditory signal processing.

Preliminary Investigation of the Passively Evoked N400 as a Tool for Estimating Speech-in-Noise Thresholds.

Speech-in-noise testing relies on a number of factors beyond the auditory system, such as cognitive function, compliance, and motor function. It may be possible to avoid these limitations by using electroencephalography. The present study explored this possibility using the N400. Eleven adults with typical hearing heard high-constraint sentences with congruent and incongruent terminal words in the presence of speech-shaped noise. Participants ignored all auditory stimulation and watched a video. The signal-to-noise ratio (SNR) was varied around each participant's behavioral threshold during electroencephalography recording. Speech was also heard in quiet. The amplitude of the N400 effect exhibited a nonlinear relationship with SNR. In the presence of background noise, amplitude decreased from high (+4 dB) to low (+1 dB) SNR but increased dramatically at threshold before decreasing again at subthreshold SNR (-2 dB). The SNR of speech in noise modulates the amplitude of the N400 effect to semantic anomalies in a nonlinear fashion. These results are the first to demonstrate modulation of the passively evoked N400 by SNR in speech-shaped noise and represent a first step toward the end goal of developing an N400-based physiological metric for speech-in-noise testing.

Speech recognition thresholds in reverberation: Effects of age and hearing loss

While young normally hearing individuals can function in moderate amounts of reverberation with a minimal reduction in speech understanding, older individuals with and without hearing loss are much more susceptible to reverberant distortions of the speech signal (Harris and Reitz, 1985; Danhauer and Johnson, 1991; Helfer, 1992). In this experiment, we evaluated the individual contributions of age and hearing loss to speech reception thresholds (SRT) for reverberant speech (T60: 0, 300, 600, and 1000 ms) in the presence of speech-shaped noise. A one-up one-down adaptive-tracking procedure was used to measure SRT in the better ear of each of 27 listeners. Results indicated, as expected, that older-hearing-impaired listeners had higher SRTs as compared to younger and older normally hearing listeners. However, the variability among listeners was very high. Multiple regression analyses predicting SRT with age and hearing loss as factors indicated that hearing loss was the most significantly contributing factor...

Cross-modal Informational Masking of Lipreading by Babble.

Whereas the energetic and informational masking effects of unintelligible babble on auditory speech recognition are well established, the present study is the first to investigate its effects on visual speech recognition. Young and older adults performed two lipreading tasks while simultaneously experiencing either quiet, speech-shaped noise, or 6-talker background babble. Both words at the end of uninformative carrier sentences and key words in everyday sentences were harder to lipread in the presence of babble than in the presence of speech-shaped noise or quiet. Contrary to the inhibitory deficit hypothesis of cognitive aging, babble had equivalent effects on young and older adults. In a follow-up experiment, neither the babble nor the speech-shaped noise stimuli interfered with performance of a face-processing task, indicating that babble selectively interferes with visual speech recognition and not with visual perception tasks per se. The present results demonstrate that babble can produce cross-modal informational masking and suggest a breakdown in audiovisual scene analysis, either because of obligatory monitoring of even uninformative speech sounds or because of obligatory efforts to integrate speech sounds even with uncorrelated mouth movements.

Abnormal intelligibility of speech in competing speech and in noise in a frequency region where audiometric thresholds are near-normal for hearing-impaired listeners

The ability to identify syllables in the presence of speech-shaped noise and a single-talker background was measured for 18 normal-hearing (NH) listeners, and for eight hearing-impaired (HI) listeners with near-normal audiometric thresholds for frequencies up to 1.5 kHz and a moderate to severe hearing loss above 2 kHz. The stimulus components were restricted to the low-frequency (≤1.5 kHz) region, where audiometric thresholds were classified clinically as normal or near normal for all listeners. Syllable identification in a speech background was measured as a function of the fundamental-frequency (F0) difference between competing voices (ranging from 1 semitone to ∼1 octave). HI listeners had poorer syllable intelligibility than NH listeners in all conditions. Intelligibility decreased by about the same amount for both groups when the F0 difference between competing voices was reduced. The results suggest that the ability to identify speech against noise or an interfering talker was disrupted in frequency regions of near-normal hearing for HI listeners, but that the ability to benefit from the tested F0 differences was not disrupted. This deficit was not predicted by the elevated absolute thresholds for speech in speech, but it was for speech in noise. It may result from supra-threshold auditory deficits associated with aging.

A Danish open-set speech corpus for competing-speech studies

Studies investigating speech-on-speech masking effects commonly use closed-set speech materials such as the coordinate response measure [Bolia et al. (2000). J. Acoust. Soc. Am. 107, 1065-1066]. However, these studies typically result in very low (i.e., negative) speech recognition thresholds (SRTs) when the competing speech signals are spatially separated. To achieve higher SRTs that correspond more closely to natural communication situations, an open-set, low-context, multi-talker speech corpus was developed. Three sets of 268 unique Danish sentences were created, and each set was recorded with one of three professional female talkers. The intelligibility of each sentence in the presence of speech-shaped noise was measured. For each talker, 200 approximately equally intelligible sentences were then selected and systematically distributed into 10 test lists. Test list homogeneity was assessed in a setup with a frontal target sentence and two concurrent masker sentences at ±50° azimuth. For a group of 16 normal-hearing listeners and a group of 15 elderly (linearly aided) hearing-impaired listeners, overall SRTs of, respectively, +1.3 dB and +6.3 dB target-to-masker ratio were obtained. The new corpus was found to be very sensitive to inter-individual differences and produced consistent results across test lists. The corpus is publicly available.

Release From Perceptual Masking for Children and Adults

The purpose of this study was to test the hypothesis that a carrier phrase can improve word recognition performance for both children and adults by providing an auditory grouping cue. It was hypothesized that the carrier phrase would benefit listeners under conditions in which they have difficulty in perceptually separating the target word from the competing background. To test this hypothesis, word recognition was examined for maskers that were believed to vary in their ability to create perceptual masking. In addition to determining the conditions under which a carrier-phrase benefit is obtained, age-related differences in both susceptibility to masking and carrier-phrase benefit were examined. Two experiments were conducted to characterize developmental effects in the ability to benefit from a carrier phrase (i.e., "say the word") before the target word. Using an open-set task, word recognition performance was measured for three listener age groups: 5- to 7-year-old children, 8- to 10-year-old children, and adults (18-30 years). For all experiments, target words were presented in each of two carrier-phrase conditions: (1) carrier-present and (2) carrier-absent. Across experiments, word recognition performance was assessed in the presence of multi-talker babble (Experiment 1), two-talker speech (Experiment 2), or speech-shaped noise (Experiment 2). Children's word recognition performance was generally poorer than that of adults for all three masker conditions. Differences between the two age groups of children were seen for both speech-shaped noise and multi-talker babble, with 5- to 7-year-olds performing more poorly than 8- to 10-year-olds. However, 5- to 7-year-olds and 8- to 10-year-olds performed similarly for the two-talker masker. Despite developmental effects in susceptibility to masking, both groups of children and adults showed a carrier-phrase benefit in multi-talker babble (Experiment 1) and in the two-talker masker (Experiment 2). The magnitude of the carrier-phrase benefit was similar for a given masker type across age groups, but the carrier-phrase benefit was greater in the presence of the two-talker masker than in multi-talker babble. Specifically, the children's average carrier-phrase benefit was 7.1% for multi-talker and 16.8% for the two-talker masker condition. No carrier-phrase benefit was observed for any age group in the presence of speech-shaped noise. Effects of auditory masking on word recognition performance were greater for children than for adults. The time course of development for susceptibility to masking seems to be more prolonged for a two-talker speech masker than for multi-talker babble or speech-shaped noise. Unique to the present study, this work suggests that a carrier phrase can provide an effective auditory grouping cue for both children and adults under conditions expected to produce substantial perceptual masking.

Performance of phonemically targeted processing in conjunction with compression processing with spectral enhancement.

The goal of the present study is to test, individually and in combination, two signal-processing strategies designed to improve both consonant and vowel perception. In the first strategy, specific consonants were targeted for processing to increase amplitude and duration. For consonants with a duration increase, the adjacent vowel was decreased proportionately in order to maintain overall word and sentence duration. Second, Coliga, an adaptive compression-processing strategy incorporating spectral enhancement, was used to process stimuli. Hearing-in-noise-test sentences were presented monaurally to normal-hearing and hearing-impaired adults in the presence of speech-shaped noise. For normal-hearing listeners, it appears that Coliga is improving intelligibility in certain conditions. Conservative amounts of duration and amplitude processing need to be applied in combination rather than when applied independently in order to provide benefit. For the hearing-impaired listeners, Coliga with spectral enhancement resulted in poorer performance than without, for all conditions except for the Coliga-only conditions, where the intelligibility was the same. It appears that phonemically-targeted speech processing and Coliga are working antagonistically due to either 1) the processing method negating the benefits when the two are combined, or 2) combining the duration and amplitude components leading to no benefit despite documented benefit of each when employed individually.

The contribution of changes in F0 and spectral tilt to increased intelligibility of speech produced in noise

Talkers modify the way they speak in the presence of noise. As well as increases in voice level and fundamental frequency (F0), a flattening of spectral tilt is observed. The resulting “Lombard speech” is typically more intelligible than speech produced in quiet, even when level differences are removed. What is the cause of the enhanced intelligibility of Lombard speech? The current study explored the relative contributions to intelligibility of changes in mean F0 and spectral tilt. The roles of F0 and spectral tilt were assessed by measuring the intelligibility gain of non-Lombard speech whose mean F0 and spectrum were manipulated, both independently and in concert, to simulate those of natural Lombard speech. In the presence of speech-shaped noise, flattening of spectral tilt contributed greatly to the intelligibility gain of noise-induced speech over speech produced in quiet while an increase in F0 did not have a significant influence. The perceptual effects of spectrum flattening was attributed to its ability of increasing the amount of speech time–frequency plane “glimpsed” in the presence of noise. However, spectral tilt changes alone could not fully account for the intelligibility of Lombard speech. Other changes observed in Lombard speech such as durational modifications may well contribute to intelligibility.