Spectral restoration of speech: Intelligibility is increased by inserting noise in spectral gaps

In order to function effectively as a means of communication, speech must be intelligible under the noisy conditions encountered in everyday life. Two types of perceptual synthesis have been reported that can reduce or cancel the effects of masking by extraneous sounds: Phonemic restoration can enhance intelligibility when segments are replaced or masked by noise, and contralateral induction can prevent mislateralization by effectively restoring speech masked at one ear when it is heard in the other. The present study reports a third type of perceptual synthesis induced by noise: enhancement of intelligibility produced by adding noise to spectral gaps. In most of the experiments, the speech stimuli consisted of two widely separated narrow bands of speech (center frequencies of 370 and 6000 Hz, each band having high-pass and low-pass slopes of 115 dB/octave meeting at the center frequency). These very narrow bands effectively reduced the available information to frequency-limited patterns of amplitude fluctuation lacking information concerning formant structure and frequency transitions. When stochastic noise was introduced into the gap separating the two speech bands, intelligibility increased for “everyday” sentences, for sentences that varied in the transitional probability of keywords, and for monosyllabic word lists. Effects produced by systematically varying noise amplitude and noise bandwidth are reported, and the implications of some of the novel effects observed are discussed.     

The effects of road traffic noise and meaningful irrelevant speech on different memory systems

To explore why noise has reliable effects on delayed recall in a certain text-reading task, this episodic memory task was employed with other memory tests in a study of road traffic noise and meaningful but irrelevant speech. Context-dependent memory was tested and self-reports of affect were taken. Participants were 96 high school students. The results showed that both road traffic noise and meaningful irrelevant speech impaired recall of the text. Retrieval in noise from semantic memory was also impaired. Attention was impaired by both noise sources, but attention did not mediate the noise effects on episodic memory. Recognition was not affected by noise. Context-dependent memory was not shown. The lack of mediation by attention, and road traffic noise being as harmful as meaningful irrelevant speech, are discussed in relation to where in the input/storing/output sequence noise has its effect and what the distinctive feature of the disturbing noise is.     

The effects of meaningful irrelevant speech and road traffic noise on teachers' attention, episodic and semantic memory

The aim of the present experiment was to examine the effects of meaningful irrelevant speech and road traffic noise on attention, episodic and semantic memory, and also to examine whether the noise effects were age-dependent. A total of 96 male and female teachers in the age range of 35-45 and 55-65 years were randomly assigned to a silent or the two noise conditions. Noise effects found in episodic memory were limited to a meaningful text, where cued recall contrary to expectations was equally impaired by the two types of noise. However, meaningful irrelevant speech also deteriorated recognition of the text, whereas road traffic noise caused no decrement. Retrieval from two word fluency tests in semantic memory showed strong effects of noise exposure, one affected by meaningful irrelevant speech and the other by road traffic noise. The results implied that both acoustic variation and the semantic interference could be of importance for noise impairments. The expected age-dependent noise effects did not show up.     

Detection of periodically and randomly interrupted noise signals

The detectability of a masked, interrupted, band-limited (150-2,400 Hz) noise signal was studied as a function of whether the interruption process was periodic or random. A standard monaural two-interval, temporal forced-choice detection procedure was employed both with and without a low-level contralateral cue (MDCC). The cue was an independent noise interrupted synchronously with the signal. Detection performance was equal under all conditions and was equivalent to the detection of an uninterrupted noise signal of equal power. Furthermore, like speech waveforms, the two types of interruption processes were discriminable only at levels at least 10–12 dB above the level required for detection.     

Recognition with and without identification: dissociative effects of meaningful encoding

In a series of four experiments, the effects of levels of processing and generation on the recognition of identified versus unidentified word fragments were examined. After studying a list of words, participants took a word fragment completion test in which half of the fragments came from studied words and half came from nonstudied words. Regardless of whether they could complete a given fragment, the participants were asked to rate the likelihood that it came from a studied word. Recognition of identified fragments was best whenever the focus of the encoding task was on meaning. Recognition of unidentified fragments did not benefit from meaningful encoding in any of the experiments reported here but did benefit from generation. It is suggested that whereas recognition with identification involves the use of meaning, recognition without identification involves the use of abstract orthographic information in memory.     

Perception of synthetic speech produced automatically by rule: Intelligibility of eight text-to-speech systems

We present the results of studies designed to measure the segmental intelligibility of eight text-to-speech systems and a natural speech control, using the Modified Rhyme Test (MRT). Results indicated that the voices tested could be grouped into four categories: natural speech, high-quality synthetic speech, moderate-quality synthetic speech, and low-quality synthetic speech. The overall performance of the best synthesis system, DECtalk-Paul, was equivalent to natural speech only in terms of performance on initial consonants. The findings are discussed in terms of recent work investigating the perception of synthetic speech under more severe conditions. Suggestions for future research on improving the quality of synthetic speech are also considered.     

Updating working memory in aircraft noise and speech noise causes different fMRI activations

The present study used fMRI/BOLD neuroimaging to investigate how visual‐verbal working memory is updated when exposed to three different background‐noise conditions: speech noise, aircraft noise and silence. The number‐updating task that was used can distinguish between “substitution processes,” which involve adding new items to the working memory representation and suppressing old items, and “exclusion processes,” which involve rejecting new items and maintaining an intact memory set. The current findings supported the findings of a previous study by showing that substitution activated the dorsolateral prefrontal cortex, the posterior medial frontal cortex and the parietal lobes, whereas exclusion activated the anterior medial frontal cortex. Moreover, the prefrontal cortex was activated more by substitution processes when exposed to background speech than when exposed to aircraft noise. These results indicate that (a) the prefrontal cortex plays a special role when task‐irrelevant materials should be denied access to working memory and (b) that, when compensating for different types of noise, either different cognitive mechanisms are involved or those cognitive mechanisms that are involved are involved to different degrees.     

On Some Counter-Examples to the Guise of the Good-Thesis: Intelligibility without Desirability

This paper argues that there are cases, which various guise of the good-theses concerning desires, intentions and actions would not allow. In these cases the agent acts for considerations that the agent does not regard as good reasons. The considerations render the actions intelligible but not desirable (where desirability and intelligibility can be objective or subjective). These cases are atypical, but nonetheless show that those guise of the good-theses which do not allow them, should be revised. In typical cases the intelligibility of desires, intentions and actions co-varies with their desirability: there are both unintelligible cases without suitable desirability characteristics and cases where desirability characteristics make the desire, intention and action intelligible. The claim here is that there are further more atypical and puzzling, but equally possible cases, where intelligibility and desirability come apart. The paper first introduces the Guise of the Good - debates about desires, intentions, and actions, and suggests distinguishing the category of “acting for a reason” from “acting for a consideration not taken to be a reason”. It then argues that while desirability entails intelligibility, and lack of intelligibility entails lack of desirability, these two cases leave conceptual room for a third category, which is that of intelligibility without desirability. This is so, whether we examine objective or subjective intelligibility and desirability. The claim is meant to apply mutatis mutandis to characteristics of desires, intentions and actions. The paper then provides possible cases of intelligibility without desirability, and defends the view against some objections.     

Perception of spasmodic dysphonia speech in background noise

The purpose of this investigation was to judge whether the Lombard effect, a characteristic change in the acoustical properties of speech produced in noise, existed in adductor spasmodic dysphonia speech, and if so, whether the effect added to or detracted from speaker intelligibility. Intelligibility, as described by Duffy, is the extent to which the acoustic signal produced by a speaker is understood by a listener based on the auditory signal alone. Four speakers with adductor spasmodic dysphonia provided speech samples consisting of low probability sentences from the Speech Perception in Noise test to use as stimuli. The speakers were first tape-recorded as they read the sentences in a quiet speaking condition and were later tape-recorded as they read the same sentences while exposed to background noise. The listeners used as subjects in this study were 50 undergraduate university students. The results of the statistical analysis indicated a significant difference between the intelligibility of the speech recorded in the quiet versus noise conditions (F(1,49) = 57.80, p < or = .001). It was concluded that a deleterious Lombard effect existed for the adductor spasmodic dysphonia speaker group, with the premise that the activation of a Lombard effect in such patients may detract from their overall speech intelligibility.     

Effects of noise on identification and serial recall of nonsense syllables in older and younger adults

The present experiment investigated the hypothesis that age-related declines in cognitive functioning are partly due to a decrease in peripheral sensory functioning. In particular, it was suggested that some of the decline in serial recall for verbal material might be due to even small amounts of degradation due to noise or hearing loss. Older and younger individuals identified and recalled nonsense syllables in order at a number of different speech-to-noise ratios. Performance on the identification task was significantly correlated with performance on a subsequent serial recall task. However, this was restricted to the case in which the stimuli were presented in a substantial amount of noise. These data show that even small changes in sensory processing can lead to real and measurable declines in cognitive functioning as measured by a serial recall task.     

Implications of evaluating and intervening with incipient stutterers

A study was conducted to investigate the factors involved in the recovery from incipient stuttering without therapeutic intervention. Fifteen subjects were seen on two occasions, once for an initial evaluation after the reported onset of stuttering, and for a second evaluation to determine if they had recovered or become chronic stutterers. Within the constraints of a small sample, several findings are presented: (1) 80% of the children recovered spontaneously from stuttering, (2) subjects who exhibited more types and frequencies of disfluencies upon onset had a poorer prognosis for recovery, (3) the age of onset, if under 4 years, had little bearing on recovery, and (4) children who were seen professionally soon after disfluency onset had the greatest prognosis for recovery. Possible implications of these results are discussed.     

A comparison of intervention for problematic speech using reinforcement with and without preferred topics

Previous research has shown that responding to the appropriate and problematic speech of individuals diagnosed with developmental disabilities with interested and uninterested listener responses, respectively, can promote more appropriate conversational engagement. However, Fisher et al. (2013) also responded to appropriate speech with access to preferred conversational topics. This study examined the influence of listener interest on the problematic speech of 8 participants and tested the additive effects of (Study 1) and participant preference for (Study 2) delivering preferred topics as reinforcement for appropriate speech. Interventions were equally effective with or without arranging access to preferred topics, but a majority of participants demonstrated a preference for intervention with contingent access to preferred topics. Caregivers and speech‐language pathologists rated the intervention procedures as acceptable and changes in participants' speech satisfactory.     

Cross-accent intelligibility of speech in noise: Long-term familiarity and short-term familiarization

Listeners must cope with a great deal of variability in the speech signal, and thus theories of speech perception must also account for variability, which comes from a number of sources, including variation between accents. It is well known that there is a processing cost when listening to speech in an accent other than one's own, but recent work has suggested that this cost is reduced when listening to a familiar accent widely represented in the media, and/or when short amounts of exposure to an accent are provided. Little is known, however, about how these factors (long-term familiarity and short-term familiarization with an accent) interact. The current study tested this interaction by playing listeners difficult-to-segment sentences in noise, before and after a familiarization period where the same sentences were heard in the clear, allowing us to manipulate short-term familiarization. Listeners were speakers of either Glasgow English or Standard Southern British English, and they listened to speech in either their own or the other accent, thereby allowing us to manipulate long-term familiarity. Results suggest that both long-term familiarity and short-term familiarization mitigate the perceptual processing costs of listening to an accent that is not one's own, but seem not to compensate for them entirely, even when the accent is widely heard in the media.     

Visual search with meaningful and non-meaningful material

Two experiments are reported which attempt to assess the effects of variations in target word, context items and instructions on performance in a visual search task. In Experiment 1, subjects were required to search through context lists of three-letter nonsense syllables (of either high or low association value) for three-letter meaningful target words (of either high or low frequency). They were given either “positive” or “negative” instructions, i.e. were told either to pick out the meaningful word or to pick out the word which was not a nonsense syllable. The results showed that visual search times were significantly influenced by both frequency of target word and association value of context items. A significant interaction was observed between type of instructions and target word frequency. The design of Experiment 2 followed that of Experiment 1, with the exceptions that nonsense syllables now became target items, and meaningful words formed the contexts. Again, nonsense syllable association value and word frequency were found to be critical in determining visual search times.     

The speech and vocalization patterns of boys with ADHD compared with boys with dyslexia and boys without learning disabilities

This research study dealt with an inquiry into the speech and vocalization patterns of boys with Attention Deficit Hyperactivity Disorder (ADHD) who were not under the influence of stimulating medication, compared with the speech and vocalization patterns of boys with reading disabilities and a control group of learners without learning disabilities. The voices of 105 participants were recorded during interviews and analyzed in the laboratory using equipment that examined the temporal speech patterns and physical features of vocalization. The speech patterns were examined with a frequency examination, speech unit length duration, and correlation between the vocalization and pauses in the speech unit. The physical features of vocalization were examined with volume and frequency scales. The research results indicated that the speech and vocalization patterns of boys with ADHD were significantly different from those of boys with reading disabilities and from the boys in the control group. The results support the assumption that speech and vocalization indicators can be used as objective indicators for the diagnosis of hyperactivity syndrome with attention and concentration difficulties.     

The Correspondence between Human Intelligibility and Physical Intelligibility: The View of Jean Ladrière

This article seeks to explain the correspondence between human intelligibility and that of the physical world by synthesizing the contributions of Jean Ladrière. Ladrière shows that the objectification function of formal symbolism in mathematics as an artificial language has operative power acquired through algorithm to represent physical reality. In physical theories, mathematics relates to observations through theoretic and empirical languages mutually interacting in a methodological circle, and nonmathematical anticipatory intention guides mathematical algorithmic exploration. Ladrière reasons that mathematics can make the physical world comprehensible because of the presence of a rational principle in both kinds of intelligibility.     

Test of Spanish sentences to measure speech intelligibility in noise conditions

This article describes the development of a test for measuring the intelligibility of speech in noise for the Spanish language, similar to the test developed by Kalikow, Stevens, and Elliot (Journal of the Acoustical Society of America, 5, 1337–1360, 1977) for the English language. The test consists of six forms, each comprising 25 high-predictability (HP) sentences and 25 low-predictability (LP) sentences. The sentences were used in a perceptual task to assess their intelligibility in babble noise across three different signal-to-noise ratio (SNR) conditions in a sample of 474 normal-hearing listeners. The results showed that the listeners obtained higher scores of intelligibility for HP sentences than for LP sentences, and the scores were lower for the higher SNRs, as was expected. The final six forms were equivalent in intelligibility and phonetic content.     

Selectivity versus summation in multiple observation tasks: evidence with spectrum parameter noise in speech

The effect of stimulus repetition was investigated in a speech identification task where intelligibility was lowered not by white noise added to the audio waveform but by ‘structural’ noise added to the spectrum parameters. Several aspects of the results argue for the improvement in intelligibility between one and two presentations being due not to statistical averaging over internal or external noise, but to increased perceptual selectivity under the influence of the first presentation's stimulus properties.