Speech Maskers Research Articles

Auditory Training Effects on the Listening Skills of Children With Auditory Processing Disorder.

Children with auditory processing disorder (APD) typically present with "listening difficulties,"' including problems understanding speech in noisy environments. The authors examined, in a group of such children, whether a 12-week computer-based auditory training program with speech material improved the perception of speech-in-noise test performance, and functional listening skills as assessed by parental and teacher listening and communication questionnaires. The authors hypothesized that after the intervention, (1) trained children would show greater improvements in speech-in-noise perception than untrained controls; (2) this improvement would correlate with improvements in observer-rated behaviors; and (3) the improvement would be maintained for at least 3 months after the end of training. This was a prospective randomized controlled trial of 39 children with normal nonverbal intelligence, ages 7 to 11 years, all diagnosed with APD. This diagnosis required a normal pure-tone audiogram and deficits in at least two clinical auditory processing tests. The APD children were randomly assigned to (1) a control group that received only the current standard treatment for children diagnosed with APD, employing various listening/educational strategies at school (N = 19); or (2) an intervention group that undertook a 3-month 5-day/week computer-based auditory training program at home, consisting of a wide variety of speech-based listening tasks with competing sounds, in addition to the current standard treatment. All 39 children were assessed for language and cognitive skills at baseline and on three outcome measures at baseline and immediate postintervention. Outcome measures were repeated 3 months postintervention in the intervention group only, to assess the sustainability of treatment effects. The outcome measures were (1) the mean speech reception threshold obtained from the four subtests of the listening in specialized noise test that assesses sentence perception in various configurations of masking speech, and in which the target speakers and test materials were unrelated to the training materials; (2) the Children's Auditory Performance Scale that assesses listening skills, completed by the children's teachers; and (3) the Clinical Evaluation of Language Fundamental-4 pragmatic profile that assesses pragmatic language use, completed by parents. All outcome measures significantly improved at immediate postintervention in the intervention group only, with effect sizes ranging from 0.76 to 1.7. Improvements in speech-in-noise performance correlated with improved scores in the Children's Auditory Performance Scale questionnaire in the trained group only. Baseline language and cognitive assessments did not predict better training outcome. Improvements in speech-in-noise performance were sustained 3 months postintervention. Broad speech-based auditory training led to improved auditory processing skills as reflected in speech-in-noise test performance and in better functional listening in real life. The observed correlation between improved functional listening with improved speech-in-noise perception in the trained group suggests that improved listening was a direct generalization of the auditory training.

周波数領域適応アルゴリズムを用いた能動音響制御による音声マスキング手法

In this paper, we propose a method to provide speech privacy by deforming the frequency characteristics of the voice. An active noise control is used in this method. Major factor determining the phonological in speech recognition is relative relationship of the feature quantity of speech called formant frequency. Breaking this relationship by control in the frequency domain, performs speech masking making it difficult to recognize phonemes. With to verify the control effect by the simulation, to examine the effectiveness of the proposed method by evaluation using quantitative indicators.

Effects of speech masking on the recognition of emotional speech: Comparison between random-splicing and reverse reproduction methods

Effects of speech masking on the recognition of emotional speech: Comparison between random-splicing and reverse reproduction methods

Extended High-Frequency Bandwidth Improves Speech Reception in the Presence of Spatially Separated Masking Speech

The hypothesis that extending the audible frequency bandwidth beyond the range currently implemented in most hearing aids can improve speech understanding was tested for normal-hearing and hearing-impaired participants using target sentences and spatially separated masking speech. The Hearing In Speech Test (HIST) speech corpus was re-recorded, and four masking talkers were recorded at a sample rate of 44.1 kHz. All talkers were male native speakers of American English. For each subject, the reception threshold for sentences (RTS) was measured in two spatial configurations. In the asymmetric configuration, the target was presented from -45° azimuth and two colocated masking talkers were presented from +45° azimuth. In the diffuse configuration, the target was presented from 0° azimuth and four masking talkers were each presented from a different azimuth: +45°, +135°, -135°, and -45°. The new speech sentences, masking materials, and configurations were presented using low-pass filter cutoff frequencies of 4, 6, 8, and 10 kHz. For the normal-hearing participants, stimuli were presented in the sound field using loudspeakers. For the hearing-impaired participants, the spatial configurations were simulated using earphones, and a multiband wide-dynamic-range compressor with a modified CAM2 fitting algorithm was used to compensate for each participant's hearing loss. For the normal-hearing participants (N = 24, mean age 40 years), the RTS improved significantly by 3.0 dB when the bandwidth was increased from 4 to 10 kHz, and a significant improvement of 1.3 dB was obtained from extending the bandwidth from 6 to 10 kHz, in both spatial configurations. Hearing-impaired participants (N = 25, mean age 71 years) also showed a significant improvement in RTS with extended bandwidth, but the effect was smaller than for the normal-hearing participants. The mean decrease in RTS when the bandwidth was increased from 4 to 10 kHz was 1.3 dB for the asymmetric condition and 0.5 dB for the diffuse condition. Extending bandwidth from 4 to 10 kHz can improve the ability of normal-hearing and hearing-impaired participants to understand target speech in the presence of spatially separated masking speech. Future studies of the benefits of extended high-frequency amplification should investigate other realistic listening situations, masker types, spatial configurations, and room reverberation conditions, to determine added value in overcoming the technical challenges associated with implementing a device capable of providing extended high-frequency amplification.

Enhancing listener strategies using a payoff matrix in speech-on-speech masking experiments.

Speech recognition was measured as a function of the target-to-masker ratio (TMR) with syntactically similar speech maskers. In the first experiment, listeners were instructed to report keywords from the target sentence. Data averaged across listeners showed a plateau in performance below 0 dB TMR when masker and target sentences were from the same talker. In this experiment, some listeners tended to report the target words at all TMRs in accordance with the instructions, while others reported keywords from the louder of the sentences, contrary to the instructions. In the second experiment, stimuli were the same as in the first experiment, but listeners were also instructed to avoid reporting the masker keywords, and a payoff matrix penalizing masker keywords and rewarding target keywords was used. In this experiment, listeners reduced the number of reported masker keywords, and increased the number of reported target keywords overall, and the average data showed a local minimum at 0 dB TMR with same-talker maskers. The best overall performance with a same-talker masker was obtained with a level difference of 9 dB, where listeners achieved near perfect performance when the target was louder, and at least 80% correct performance when the target was the quieter of the two sentences.

Delayed Stream Segregation in Older Adults: More Than Just Informational Masking.

To determine whether the time course for the buildup of auditory stream segregation differs between younger and older adults. Word recognition thresholds were determined for the first and last keywords in semantically anomalous but syntactically correct sentences (e.g., "A rose could paint a fish") when the target sentences were masked by speech-spectrum noise, 3-band vocoded speech, 16-band vocoded speech, intact and colocated speech, and intact and spatially separated speech. A significant reduction in thresholds from the first to the last keyword was interpreted as indicating that stream segregation improved with time. The buildup of stream segregation is slowed for both age groups when the masker is intact, colocated speech. Older adults are more disadvantaged; for them, stream segregation is also slowed even when a speech masker is spatially separated, conveys little meaning (3-band vocoding), and vocal fine structure cues are impoverished but envelope cues remain available (16-band vocoding).

Benefits of Acoustic Beamforming for Solving the Cocktail Party Problem.

The benefit provided to listeners with sensorineural hearing loss (SNHL) by an acoustic beamforming microphone array was determined in a speech-on-speech masking experiment. Normal-hearing controls were tested as well. For the SNHL listeners, prescription-determined gain was applied to the stimuli, and performance using the beamformer was compared with that obtained using bilateral amplification. The listener identified speech from a target talker located straight ahead (0° azimuth) in the presence of four competing talkers that were either colocated with, or spatially separated from, the target. The stimuli were spatialized using measured impulse responses and presented via earphones. In the spatially separated masker conditions, the four maskers were arranged symmetrically around the target at ±15° and ±30° or at ±45° and ±90°. Results revealed that masked speech reception thresholds for spatially separated maskers were higher (poorer) on average for the SNHL than for the normal-hearing listeners. For most SNHL listeners in the wider masker separation condition, lower thresholds were obtained through the microphone array than through bilateral amplification. Large intersubject differences were found in both listener groups. The best masked speech reception thresholds overall were found for a hybrid condition that combined natural and beamforming listening in order to preserve localization for broadband sources.

Musical training, individual differences and the cocktail party problem.

Are musicians better able to understand speech in noise than non-musicians? Recent findings have produced contradictory results. Here we addressed this question by asking musicians and non-musicians to understand target sentences masked by other sentences presented from different spatial locations, the classical ‘cocktail party problem’ in speech science. We found that musicians obtained a substantial benefit in this situation, with thresholds ~6 dB better than non-musicians. Large individual differences in performance were noted particularly for the non-musically trained group. Furthermore, in different conditions we manipulated the spatial location and intelligibility of the masking sentences, thus changing the amount of ‘informational masking’ (IM) while keeping the amount of ‘energetic masking’ (EM) relatively constant. When the maskers were unintelligible and spatially separated from the target (low in IM), musicians and non-musicians performed comparably. These results suggest that the characteristics of speech maskers and the amount of IM can influence the magnitude of the differences found between musicians and non-musicians in multiple-talker “cocktail party” environments. Furthermore, considering the task in terms of the EM-IM distinction provides a conceptual framework for future behavioral and neuroscientific studies which explore the underlying sensory and cognitive mechanisms contributing to enhanced “speech-in-noise” perception by musicians.

Informational masking of speech in dyslexic children.

Studies evaluating speech perception in noise have reported inconsistent results regarding a potential deficit in dyslexic children. So far, most of them investigated energetic masking. The present study evaluated situations inducing mostly informational masking, which reflects cognitive interference induced by the masker. Dyslexic children were asked to identify a female target syllable presented in quiet, babble, unmodulated, and modulated speech-shaped noise. Whereas their performance was comparable to normal-reading children in quiet, it dropped significantly in all noisy conditions compared to age-, but not reading level-matched controls. Interestingly, noise affected similarly the reception of voicing, place, and manner of articulation in dyslexic and normal-reading children.

Discrimination of Stochastic Frequency Modulation by Cochlear Implant Users.

Past work has shown that low-rate frequency modulation (FM) may help preserve signal coherence, aid segmentation at word and syllable boundaries, and benefit speech intelligibility in the presence of a masker. This study evaluated whether difficulties in speech perception by cochlear implant (CI) users relate to a deficit in the ability to discriminate among stochastic low-rate patterns of FM. This is a correlational study assessing the association between the ability to discriminate stochastic patterns of low-rate FM and the intelligibility of speech in noise. Thirteen postlingually deafened adult CI users participated in this study. Using modulators derived from 5-Hz lowpass noise applied to a 1-kHz carrier, thresholds were measured in terms of frequency excursion both in quiet and with a speech-babble masker present, stimulus duration, and signal-to-noise ratio in the presence of a speech-babble masker. Speech perception ability was assessed in the presence of the same speech-babble masker. Relationships were evaluated with Pearson product-moment correlation analysis with correction for family-wise error, and commonality analysis to determine the unique and common contributions across psychoacoustic variables to the association with speech ability. Significant correlations were obtained between masked speech intelligibility and three metrics of FM discrimination involving either signal-to-noise ratio or stimulus duration, with shared variance among the three measures accounting for much of the effect. Compared to past results from young normal-hearing adults and older adults with either normal hearing or a mild-to-moderate hearing loss, mean FM discrimination thresholds obtained from CI users were higher in all conditions. The ability to process the pattern of frequency excursions of stochastic FM may, in part, have a common basis with speech perception in noise. Discrimination of differences in the temporally distributed place coding of the stimulus could serve as this common basis for CI users.

Informational masking of monaural target speech by a single contralateral formant.

Recent research suggests that the ability of an extraneous formant to impair intelligibility depends on the variation of its frequency contour. This idea was explored using a method that ensures interference cannot occur through energetic masking. Three-formant (F1 + F2 + F3) analogues of natural sentences were synthesized using a monotonous periodic source. Target formants were presented monaurally, with the target ear assigned randomly on each trial. A competitor for F2 (F2C) was presented contralaterally; listeners must reject F2C to optimize recognition. In experiment 1, F2Cs with various frequency and amplitude contours were used. F2Cs with time-varying frequency contours were effective competitors; constant-frequency F2Cs had far less impact. To a lesser extent, amplitude contour also influenced competitor impact; this effect was additive. In experiment 2, F2Cs were created by inverting the F2 frequency contour about its geometric mean and varying its depth of variation over a range from constant to twice the original (0%-200%). The impact on intelligibility was least for constant F2Cs and increased up to ∼100% depth, but little thereafter. The effect of an extraneous formant depends primarily on its frequency contour; interference increases as the depth of variation is increased until the range exceeds that typical for F2 in natural speech.

Acoustic detail guides attention allocation in a selective listening task.

The flexible allocation of attention enables us to perceive and behave successfully despite irrelevant distractors. How do acoustic challenges influence this allocation of attention, and to what extent is this ability preserved in normally aging listeners? Younger and healthy older participants performed a masked auditory number comparison while EEG was recorded. To vary selective attention demands, we manipulated perceptual separability of spoken digits from a masking talker by varying acoustic detail (temporal fine structure). Listening conditions were adjusted individually to equalize stimulus audibility as well as the overall level of performance across participants. Accuracy increased, and response times decreased with more acoustic detail. The decrease in response times with more acoustic detail was stronger in the group of older participants. The onset of the distracting speech masker triggered a prominent contingent negative variation (CNV) in the EEG. Notably, CNV magnitude decreased parametrically with increasing acoustic detail in both age groups. Within identical levels of acoustic detail, larger CNV magnitude was associated with improved accuracy. Across age groups, neuropsychological markers further linked early CNV magnitude directly to individual attentional capacity. Results demonstrate for the first time that, in a demanding listening task, instantaneous acoustic conditions guide the allocation of attention. Second, such basic neural mechanisms of preparatory attention allocation seem preserved in healthy aging, despite impending sensory decline.

Need of speech-in-noise testing to assess listening difficulties in older adults

Objective: To assess the need of including speech-in-noise tests in the audiological test battery in older adults and the key factors that affect performance in these tests. Older adults complain significantly more than other age groups regarding difficulties in hearing speech in noise, even in the presence of a normal audiogram. Hearing speech in noise is a very complex function underpinned by high-level, cognitive and analytic processing along with low-level sensory processing. Study design: The design reflected an overview of the effect of aging in the auditory system and the acoustic and cognitive resources required for hearing speech in noise. In addition, a review of key clinical studies of speech-in-noise tests was conducted in older adults. Results: Age affects the peripheral and central auditory system but also the entire brain at a structural, functional and neurochemical level. Listening to speech in noise requires complex and multimodal processing in auditory, linguistic, cognitive and other domains via multiple brain interconnections that are not yet fully understood. There are several different theories and models proposed to explain age related declines in speech-in-noise performance. Listeners may give different scores in speech-in-noise tests depending on the parameters of these tests, e.g. type of speech stimulus (words vs. sentences) and masker (noise, multi-talker babbler). Speech recognition in noise depends on the degree of direct interaction of the target and masker at the cochlear level (energetic masking) and on how well speech can be dissociated from the masker (informational masking). Studies show more pronounced effects of informational masking and of lexical interference in older adults, who also show poorer use of auditory sensory type cues, and require greater listening effort in order to perform speech-in-noise tasks. Conclusion: Assessing older adults with pure tone audiometry only and addressing these difficulties with hearing amplification may not be enough to address older adults’ hearing needs, as this approach does not take into account the higher level impairments contributing to hearing and listening problems. Careful audiological assessment with appropriately chosen speech-in-noise tests, together with audiometry and potentially additional tests, are thus required in order to identify and characterize hearing difficulties in older adults, in order to inform their management plan.

Masked speech recognition in a single-talker or a single-talker-plus-noise masker in school-age children and adults

Previous work has shown that masked speech perception is poorer in younger children than older children and adults, particularly when the masker is composed of speech. The present study tested the hypothesis that young children’s poor performance in speech maskers is due to reduced ability to benefit from transient improvements in the target-to-masker ratio (TMR). Listeners were normal-hearing school-age children (5–15 years) and adults. The target words and masker speech were produced by two different female talkers. Target stimuli were disyllabic words, and the masker was either: (1) speech-only or (2) speech-plus-noise. Both maskers included a continuous 60-dB-SPL stream of speech. The speech-plus-noise masker also included a continuous 50-dB-SPL speech-shaped noise. A four-alternative forced choice task was used, and target level was adapted to estimate threshold. Thresholds were higher for younger children than older children or adults for both masker conditions. While the inclusion of speech-shaped noise increased thresholds for all listeners, it had a smaller detrimental effect on younger children than on older children and adults. This result is consistent with the idea that greater susceptibility to speech-on-speech masking in younger children is related to a reduced ability to benefit from transient improvements in TMR.

Informational masking of monaural speech by a single contralateral formant

Recent research suggests that the ability of an extraneous formant to impair intelligibility depends on the variation of its frequency contour. This idea was explored using a method that ensures interference occurs only through informational masking. Three-formant analogues of sentences were synthesized using a monotonous periodic source (F0 = 140 Hz). Target formants were presented monaurally; the target ear was assigned randomly on each trial. A competitor for F2 (F2C) was presented contralaterally; listeners must reject F2C to optimize recognition. In experiment 1, F2Cs with various frequency and amplitude contours were used. F2Cs with time-varying frequency contours were effective competitors; constant-frequency F2Cs had far less impact. Amplitude contour also influenced competitor impact; this effect was additive. In experiment 2, F2Cs were created by inverting the F2 frequency contour about its geometric mean and varying its depth of variation over a range from constant to twice the original (0–200%). The impact on intelligibility was least for constant F2Cs and increased up to ~100% depth, but little thereafter. The effect of an extraneous formant depends primarily on its frequency contour; interference increases as the depth of variation is increased until the range exceeds that typical for F2 in natural speech. [Work supported by ESRC.]

Application of a monaural glimpsing model to binaural speech mixtures

Under certain circumstances, listeners with sensorineural hearing loss demonstrate poorer speech intelligibility in spatially separated speech maskers than those with normal hearing. One important issue in interpreting these results is whether the target speech information is available or audible in the spatialized mixture. Simple energy-based “glimpsing” models can be used to quantify the available target in speech mixtures. For example, as target level decreases in a fixed-level masker, fewer and fewer good glimpses are available. Moreover, those glimpses may not all be above threshold, particularly for individuals with hearing loss. In this study we used a glimpsing model to isolate available target glimpses in binaural speech stimuli (separately for each ear) as a function of target-to-masker ratio. Performance for stimuli processed in this way was compared to binaural performance in listeners with normal hearing and listeners with hearing loss. Results suggest that binaural performance was partly limited by the ability to use the available glimpses. Increasing the level of the glimpses gave mixed results, which will be discussed in terms of audibility. [Work supported by NIH/NIDCD.]

Adult listener’s detection thresholds for vowels and two-syllable words in different two-talker maskers

There is an increased interest in using natural speech targets and speech maskers to assess infants’ auditory skill development. While providing ecological validity, natural speech is an inherently uncontrolled signal. Discrepancies between existing studies may be the result of differences in the maskers and/or target stimuli used. This study explores adult listener’s detection thresholds for vowels and two-syllable words in three different two-talker maskers to assess the impact of (1) masker characteristics and (2) differences in target signals. Each two-talker masker was created by recording two female speakers reading different passages in monotone, in adult-directed style, or infant-directed style. Silences longer than 300 ms were deleted, and the passages’ root-mean-square amplitudes balanced before combining them. The target stimuli consisted of the vowels /a/, /o/, /e/ and the words “baby” and “ice-cream.” They were spoken by different female talkers in both adult-directed and infant-directed styles. An adaptive two-interval forced choice method was used to estimate threshold. Preliminary results suggest that there are small differences in detection thresholds between the different two-talker maskers. Larger differences are found between the thresholds of different target signals, indicating that discrepancies between existing studies may be due to differences in target stimuli used.

Selective spatial attention modulates bottom-up informational masking of speech

To hear out a conversation against other talkers listeners overcome energetic and informational masking. Largely attributed to top-down processes, information masking has also been demonstrated using unintelligible speech and amplitude-modulated maskers suggesting bottom-up processes. We examined the role of speech-like amplitude modulations in information masking using a spatial masking release paradigm. Separating a target talker from two masker talkers produced a 20 dB improvement in speech reception threshold; 40% of which was attributed to a release from informational masking. When across frequency temporal modulations in the masker talkers are decorrelated the speech is unintelligible, although the within frequency modulation characteristics remains identical. Used as a masker as above, the information masking accounted for 37% of the spatial unmasking seen with this masker. This unintelligible and highly differentiable masker is unlikely to involve top-down processes. These data provides strong evidence of bottom-up masking involving speech-like, within-frequency modulations and that this, presumably low level process, can be modulated by selective spatial attention.

Subjective and objective rating of spectrally different pseudorandom noises--implications for speech masking design.

Artificial sound masking is increasingly used in open-plan offices to improve speech privacy and to reduce distraction caused by speech sounds. Most of the masking sounds are based on pseudorandom continuous noise filtered to a specific spectrum that should be optimized in respect with speech masking efficiency and comfort. The aim of this study was to increase basic understanding regarding the comfort. The second aim was to determine how well objective rating methods (15 different noise indices) predict the subjective ratings. Twenty-three subjects rated the loudness, disturbance, pleasantness, and six other subjective measures of 11 spectrally different noises in laboratory conditions. Speech was not present during the experiment. All sounds were presented at 42 dB LAeq within 50-10,000 Hz. Unexpectedly, the subjects were most satisfied with sounds having emphasis on low frequencies. A sound having a slope of -7 dB per octave increment resulted in the highest satisfaction. Changes in subjective ratings were reasonably well predicted by five noise indices, while many well-known noise indices frequently used in building design underperformed in this task. The results are expected to benefit in the design of masking sounds and other appliances.

The effect of spatial separation in distance on the intelligibility of speech in rooms.

The influence of spatial separation in source distance on speech reception thresholds (SRTs) is investigated. In one scenario, the target was presented at 0.5 m distance, and the masker varied from 0.5 m distance up to 10 m. In a second scenario, the masker was presented at 0.5 m distance and the target distance varied. The stimuli were synthesized using convolution with binaural room impulse responses (BRIRs) measured on a dummy head in a reverberant auditorium, and were equalized to compensate for distance-dependent spectral and intensity changes. All sources were simulated directly in front of the listener. SRTs decreased monotonically when the target was at 0.5 m and the speech-masker was moved further away, resulting in a SRT improvement of up to 10 dB. When the speech masker was at 0.5 m and the target was moved away, a large variation across subjects was observed. Neither short-term signal-to-noise ratio (SNR) improvements nor cross-ear glimpsing could account for the observed improvement in intelligibility. However, the effect might be explained by an improvement in the SNR in the modulation domain and a decrease in informational masking. This study demonstrates that distance-related cues can play a significant role when listening in complex environments.