Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “SPEECH”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 73 records · Page 4Linked to original sources
PubMed2005

Statistical properties of infant-directed versus adult-directed speech: insights from speech recognition.

Previous studies have shown that infant-directed speech ('motherese') exhibits overemphasized acoustic properties which may facilitate the acquisition of phonetic categories by infant learners. It has been suggested that the use of infant-directed data for training automatic speech recognition systems might also enhance the automatic learning and discrimination of phonetic categories. This study investigates the properties of infant-directed vs. adult-directed speech from the point of view of the statistical pattern recognition paradigm underlying automatic speech recognition. Isolated-word speech recognizers were trained on adult-directed vs. infant-directed data sets and were tested on both matched and mismatched data. Results show that recognizers trained on infant-directed speech did not always exhibit better recognition performance; however, their relative loss in performance on mismatched data was significantly less severe than that of recognizers trained on adult-directed speech and presented with infant-directed test data. An analysis of the statistical distributions of a subset of phonetic classes in both data sets showed that this pattern is caused by larger class overlaps in infant-directed speech. This finding has implications for both automatic speech recognition and theories of infant speech perception.

Adult↗
PubMed1992

Speech pattern hearing aids for the profoundly hearing impaired: speech perception and auditory abilities.

A family of prototype speech pattern hearing aids for the profoundly hearing impaired has been compared to amplification. These aids are designed to extract acoustic speech patterns that convey essential phonetic contrasts, and to match this information to residual receptive abilities. In the first study, the presentation of voice fundamental frequency information from a wearable SiVo (sinusoidal voice) aid was compared to amplification in 11 profoundly deafened adults. Intonation reception was often better, and never worse, with fundamental frequency information. Four subjects scored more highly in audio-visual consonant identification with fundamental frequency information, five performed better with amplified speech, and two performed similarly under these two conditions. Five of the 11 subjects continued use of the SiVo aid after the tests were complete. A second study examined a laboratory prototype compound speech pattern aid, which encoded voice fundamental frequency, amplitude envelope, and the presence of voiceless excitation. In five profoundly deafened adults, performance was better in consonant identification when additional speech patterns were present than with fundamental frequency alone; the main advantage was derived from amplitude information. In both consonant identification and connected discourse tracking, performance with appropriately matched compound speech pattern signals was better than with amplified speech in three subjects, and similar to performance with amplified speech in the other two. In nine subjects, frequency discrimination, gap detection, and frequency selectivity were measured, and were compared to speech receptive abilities with both amplification and fundamental frequency presentation. The subjects who showed the greatest advantage from fundamental frequency presentation showed the greatest average hearing losses, and the least degree of frequency selectivity. Compound speech pattern aids appear to be more effective for some profoundly hearing-impaired listeners than conventional amplifying aids, and may be a valuable alternative to cochlear implants.

Adult↗
PubMed2005

Discrimination and categorization of speech and non-speech sounds in an MEG delayed-match-to-sample study.

We investigated the perception and categorization of speech (vowels, syllables) and non-speech (tones, tonal contours) stimuli using MEG. In a delayed-match-to-sample paradigm, participants listened to two sounds and decided if they sounded exactly the same or different (auditory discrimination, AUD), or if they belonged to the same or different categories (category discrimination, CAT). Stimuli across the two conditions were identical; the category definitions for each kind of sound were learned in a training session before recording. MEG data were analyzed using an induced wavelet transform method to investigate task-related differences in time-frequency patterns. In auditory cortex, for both AUD and CAT conditions, an alpha (8-13 Hz) band activation enhancement during the delay period was found for all stimulus types. A clear difference between AUD and CAT conditions was observed for the non-speech stimuli in auditory areas and for both speech and non-speech stimuli in frontal areas. The results suggest that alpha band activation in auditory areas is related to both working memory and categorization for new non-speech stimuli. The fact that the dissociation between speech and non-speech occurred in auditory areas, but not frontal areas, points to different categorization mechanisms and networks for newly learned (non-speech) and natural (speech) categories.

Acoustic Stimulation↗
PubMed1991

The effects of digital quantization error on speech intelligibility and perceived speech quality.

The effects of digital quantization error upon speech intelligibility and perceived speech quality, for normally hearing subjects, were investigated for digitized speech processed to simulate 6-, 8-, 10-, 12-, 14-, and 16-bit integer conversion and 2-, 3-, 4-, 5-, 6-, and 7-bit floating-point conversion. For the integer data, there were no significant differences in speech intelligibility for 8- to 16-bit conversion. Only 6-bit integer conversion at 55 dB SPL resulted in a significant degradation in speech intelligibility. For the floating-point data, there were no significant differences in speech intelligibility for 2- to 7-bit floating-point conversion. However, results of the perceived quality experiment appeared to be more sensitive to differences among the various conditions. Speech processed using 12-, 14-, and 16-bit integer conversion was judged to be superior to speech processed using the 6-, 8-, and 10-bit integer conditions. Speech processed using 5-, 6-, and 7-bit floating-point conversion was judged to be superior to speech processed using 2-, 3-, and 4-bit floating-point conversion.

Acoustics↗
PubMed1996

Speech timing in apraxia of speech versus conduction aphasia.

This study examined temporal parameters of speech in subjects with apraxia of speech, conduction aphasia, and normal speech. They were asked to repeat target words in a carrier phrase 10 times. Acoustic analyses involved measurement of stop gap duration, voice onset time, vowel nucleus duration, and consonant-vowel (CV) duration. Speakers with apraxia of speech had longer and more variable stop gap, vowel, and CV durations than did subjects with aphasia or normal speech. Speakers with conduction aphasia had longer vowel durations and CV durations than subjects with normal speech. Also, subjects with apraxia of speech showed greater token-to-token variability than the other subject groups. The variability shown by subjects with apraxia of speech was significantly correlated with perceptual judgments of their speech. The significance of these results is discussed in the context of motoric and phonological explanations for apraxia of speech and conduction aphasia.

Adult↗
PubMed2006

Auditory-visual speech perception and synchrony detection for speech and nonspeech signals.

Previous research has identified a "synchrony window" of several hundred milliseconds over which auditory-visual (AV) asynchronies are not reliably perceived. Individual variability in the size of this AV synchrony window has been linked with variability in AV speech perception measures, but it was not clear whether AV speech perception measures are related to synchrony detection for speech only or for both speech and nonspeech signals. An experiment was conducted to investigate the relationship between measures of AV speech perception and AV synchrony detection for speech and nonspeech signals. Variability in AV synchrony detection for both speech and nonspeech signals was found to be related to variability in measures of auditory-only (A-only) and AV speech perception, suggesting that temporal processing for both speech and nonspeech signals must be taken into account in explaining variability in A-only and multisensory speech perception.

Acoustic Stimulation↗
PubMed2005

Correlation between acoustic speech characteristics and non-speech motor performance in Parkinson Disease.

BACKGROUND: Research has found that non-speech motor deficits in Parkinson Disease (PD) can be consistently improved by levodopa medications and surgical treatment, but that speech and voice are only partially responsive to treatment. This has led to the assertion that speech is an axial (non-peripheral) feature of PD, along with other features that are only partially responsive to treatment (e.g., postural stability and gait). The current study tested this assertion via examination of the relationship between multiple measures of speech production and multiple measures of non-speech movement in individuals with PD. MATERIAL/METHODS: Nine individuals with idiopathic PD were studied, and all participants were taking levodopa-carbidopa medication. Motor performance was examined using the motor portion of the Unified Parkinson's Disease Rating Scale, and speech was examined via acoustic analysis of phonation, articulation, and prosody. RESULTS: Seven of sixteen speech acoustic measures (FoSD in vowels, F2 slope for /u/ and /ae/, FoSD in reading, articulation rate in monologue, and percent pause in reading and monologue) were significantly correlated with non-speech movements. Results suggested that speech measures are correlated with both axial motor symptoms (e.g., gait, facial expression, posture, postural stability) and non-axial motor symptoms (e.g., rest tremor, left and right bradykinesia, postural tremor). CONCLUSIONS: It has been hypothesized that axial symptoms of PD are more purely dopaminergic, and non-axial symptoms are related to non-dopaminergic lesions. Therefore, the current results indicate that certain speech deficits in PD may result from dopaminergic lesions, while others appear to result from non-dopaminergic lesions.

Aged↗
PubMed1997

[Progressive loss of speech output with festinating speech--a case report].

Primary progressive aphasia is a rare disorder of unknown cause. We report a patient with progressive loss of speech output, a clinical variant of PPA, characterized by festinating speech. A 60-year-old right handed woman was admitted to our hospital, because of progressive deterioration of her speech. On admission, she was alert and orientated without dementia. A severe impairment of her articulation was observed: her speech rate was so fast that her speech became almost intelligible. The orofacial apraxia and difficulty in tapping were also present. The other neurological findings were normal. Neuroradiological studies showed the left perisylvian atrophy. Festinating speech has not been previously reported in patients with PPA; patient with PPA usually show a slow speech rate with effortful expression. Since festinating speech is occasionally present in the extrapyramidal disorders, such as Parkinson's disease, progressive supranuclear palsy, or pure akinesia, it appears likely that the combined lesions of the perisylvian region and the basal ganglia are responsible for her characteristic speech disorder with festinating speech.

Aphasia↗
PubMed1989

The Indiana Speech Training Aid (ISTRA). I: Comparisons between human and computer-based evaluation of speech quality.

Experimental comparisons are reported between computer-based and human judgments of speech quality for the same sets of utterances. Speech stimuli were recorded from two normal talkers, who intentionally varied the quality of their speech, and from a hearing-impaired child who was receiving speech therapy on the Indiana Speech Training Aid (ISTRA). The tape recordings were submitted for evaluation to a naive jury, an expert jury, and the ISTRA System, a microcomputer equipped with a speaker-dependent speech recognition board that generated scores representing how well utterance matched a stored template. Correlational analyses of these data indicated that humans were slightly better at judging speech quality than was the computer, but that the computer was much more reliable. These results demonstrate that computer-based speech evaluation may be a reasonable substitute for human judgments for certain types of speech drill.

Computer Systems↗
PubMed1993

Segmental intelligibility and speech interference thresholds of high-quality synthetic speech in presence of noise.

Technological advancement in the area of synthetic speech has made it increasingly difficult to distinguish quality of speech based solely on intelligibility scores obtained in benign laboratory conditions. Intelligibility scores obtained for natural speech and a high-quality text-to-speech system (DECtalk) are not substantially different. This study examined the perceived intelligibility and speech interference thresholds of DECtalk male and female voices and compared them with data obtained for natural speech. Results revealed that decreasing signal-to-noise levels had more deleterious effects on the perception of DECtalk male and female voices than on the perception of natural speech. Analysis of pattern of phoneme errors revealed that similar general patterns of errors tended to occur in DECtalk and in natural speech. The speech interference test did not demonstrate any significant difference between the DECtalk male and female voices. These results were supported by the absence of a significant difference between DECtalk male and female voices during intelligibility testing at different signal-to-noise ratios.

Acoustic Stimulation↗
PubMedDate not supplied

Effects of speech therapy and pharmacologic and surgical treatments on voice and speech in Parkinson's disease: a review of the literature.

The purpose of this review was to examine the different treatment approaches for persons with Parkinson's Disease (PD) and to examine the effects of these treatments on speech. Treatment methods reviewed include speech therapy, pharmacological, and surgical. Research from the 1950s through the 1970s had not demonstrated significant improvements following speech therapy. Recent research has shown that speech therapy (when persons with PD are optimally medicated) has proven to be the most efficacious therapeutic method for improving voice and speech function. Pharmacological methods of treatment in isolation do not appear to significantly improve voice and speech function in PD across research studies. Surgical treatment methods including pallidotomy and deep brain stimulation may be significant treatment options which improve voice and speech function in some persons with PD. Possible explanations for the differential responses to treatment are discussed. Future studies should investigate the effects of combined treatment approaches. Perhaps the combination of pharmacological, surgical and speech treatment will prove superior to treatments combining pharmacological and surgical or pharmacological and speech therapy in improving the communication abilities of persons with PD.

Antiparkinson Agents↗
PubMed2006

Isolating the energetic component of speech-on-speech masking with ideal time-frequency segregation.

When a target speech signal is obscured by an interfering speech wave form, comprehension of the target message depends both on the successful detection of the energy from the target speech wave form and on the successful extraction and recognition of the spectro-temporal energy pattern of the target out of a background of acoustically similar masker sounds. This study attempted to isolate the effects that energetic masking, defined as the loss of detectable target information due to the spectral overlap of the target and masking signals, has on multitalker speech perception. This was achieved through the use of ideal time-frequency binary masks that retained those spectro-temporal regions of the acoustic mixture that were dominated by the target speech but eliminated those regions that were dominated by the interfering speech. The results suggest that energetic masking plays a relatively small role in the overall masking that occurs when speech is masked by interfering speech but a much more significant role when speech is masked by interfering noise.

Adolescent↗
PubMed1998

Evaluation of the effect of speech-rate slowing on speech intelligibility in noise using a simulation of cochlear hearing loss.

The effect of digital processing, which slows the speed of speech (speech-rate) without changing its pitch, has been examined. The processing is intended to make speech communication easier by allowing more time for cognitive processing when the listening situation is difficult, for example, when listening to a foreign language, or when the user has a hearing loss. The speech-rate slowing makes use of a pitch-synchronous partial expansion of the waveform in the time domain. The processing was evaluated using a simulation of hearing loss which has been shown to lead to reduced intelligibility for normally hearing subjects. The simulation included the major consequences of cochlear hearing loss; loudness recruitment, threshold elevation, and reduced frequency selectivity. Two simulations were used: a moderate flat hearing loss with auditory filters broadened by a constant factor of three (B3R2); and the same loss with linear amplification applied prior to the simulation processing (B3R2+). Two expansion rates were used for the speech-rate slowing, 1.25 and 1.50. The intelligibility of sentences in speech-shaped noise was measured. For both simulation conditions, the speech-rate slowing did not give any improvement in intelligibility. Rather, in condition B3R2+ the slowing produced statistically significant deleterious effects on intelligibility. The results suggest that artificial speech-rate slowing will not improve the intelligibility of speech in noise for hearing-impaired people who have the type of cochlear damage simulated in this test.

Adult↗
PubMedDate not supplied

Use of the speech pattern audiometer and the electropalatograph to explore the speech production/perception relationship in a profoundly deaf child.

The use of electropalatography (EPG) in the teaching of new speech skills to deaf speakers, is not widely researched. Whether these skills can be maintained, or whether they can become fully automatic, without enough speech perception to enable auditory self-monitoring is therefore unclear. Most approaches to increasing speech intelligibility in deaf speakers, rightly, place an emphasis on maximizing residual hearing and listening skills. However, a small amount of evidence exists which suggests that speech production can aid speech perception. This research aimed to measure change in speech perception, through use of the speech pattern audiometer (SPA), whilst working on speech production with the electropalatograph. Intervention took place over a year. The issues of maintenance and generalization are discussed.

Audiometry, Speech↗
PubMed2006

Speech perception and short-term memory deficits in persistent developmental speech disorder.

Children with developmental speech disorders may have additional deficits in speech perception and/or short-term memory. To determine whether these are only transient developmental delays that can accompany the disorder in childhood or persist as part of the speech disorder, adults with a persistent familial speech disorder were tested on speech perception and short-term memory. Nine adults with a persistent familial developmental speech disorder without language impairment were compared with 20 controls on tasks requiring the discrimination of fine acoustic cues for word identification and on measures of verbal and nonverbal short-term memory. Significant group differences were found in the slopes of the discrimination curves for first formant transitions for word identification with stop gaps of 40 and 20 ms with effect sizes of 1.60 and 1.56. Significant group differences also occurred on tests of nonverbal rhythm and tonal memory, and verbal short-term memory with effect sizes of 2.38, 1.56, and 1.73. No group differences occurred in the use of stop gap durations for word identification. Because frequency-based speech perception and short-term verbal and nonverbal memory deficits both persisted into adulthood in the speech-impaired adults, these deficits may be involved in the persistence of speech disorders without language impairment.

Adolescent↗
PubMed1999

A method to determine the speech transmission index from speech waveforms.

A method for computing the speech transmission index (STI) using real speech stimuli is presented and evaluated. The method reduces the effects of some of the artifacts that can be encountered when speech waveforms are used as probe stimuli. Speech-based STIs are computed for conversational and clearly articulated speech in several noisy, reverberant, and noisy-reverberant environments and compared with speech intelligibility scores. The results indicate that, for each speaking style, the speech-based STI values are monotonically related to intelligibility scores for the degraded speech conditions tested. Therefore, the STI can be computed using speech probe waveforms and the values of the resulting indices are as good predictors of intelligibility scores as those derived from MTFs by theoretical methods.

Electronic Data Processing↗
PubMed2002

Distinct gamma-band evoked responses to speech and non-speech sounds in humans.

To understand spoken language, the human brain must have fast mechanisms for the representation and identification of speech sounds. Stimulus-induced synchronization of neural activity at gamma frequencies (20-80 Hz), occurring in humans at 200-300 msec from stimulus onset, has been suggested to be a possible mechanism for neural object representation. Auditory and visual stimuli also evoke an earlier (peak <100 msec) gamma oscillation, but its dependence on high-level stimulus parameters and, thereby, its involvement in object representation has remained unclear. Using whole-scalp magnetoencephalography, we show here that responses evoked by speech and non-speech sounds differed in the gamma-frequency but not in the low-frequency (0.1-20 Hz) band as early as 40-60 msec from stimulus onset. The gamma-band responses to the speech sound peaked earlier in the left than in the right hemisphere, whereas those to the non-speech sound peaked earlier in the right hemisphere. For the speech sound, there was no difference in the response amplitude between the hemispheres at low (20-45 Hz) gamma frequencies, whereas for the non-speech sound, the amplitude was larger in the right hemisphere. These results suggest that evoked gamma-band activity may indeed be sensitive to high-level stimulus properties and may hence reflect the neural representation of speech sounds. Consequently, speech-specific neuronal processing may commence no later than 40-60 msec from stimulus onset, possibly in the form of activation of language-specific memory traces.

Acoustic Stimulation↗
PubMed2006

Spontaneous speech events in two speech databases of human-computer and human-human dialogs in Spanish.

Previous works in English have revealed that disfluencies follow regular patterns and that incorporating them into the language model of a speech recognizer leads to lower perplexities and sometimes to a better performance. Although work on disfluency modeling has been applied outside the English community (e.g., in Japanese), as far as we know there is no specific work dealing with disfluencies in Spanish. In this paper, we follow a data driven approach in exploring the potential benefit of modeling disfluencies in a speech recognizer in Spanish. Two databases of human-computer and human-human dialogs are considered, which allow the absolute and relative frequencies of disfluencies in the two situations to be compared. The rate of disfluencies in human-human dialogs is found to be very close to that found for similar databases in English. Due to setup factors, the rate of disfluencies found in human-computer dialogs was remarkably higher than that reported for similar databases in English. In any case, from the point of view of speech recognition, the high frequencies of disfluencies and the distinct features of the acoustic events related to them support the need for explicit acoustic models. The regularities observed in the distribution of filled pauses and speech repairs reveal that including them in the language model of the speech recognizer may be also helpful. The extent to which the number of events depends on utterance length and on the speaker is also explored. Statistics are shown that follow previous studies for English, and a sizeable space is devoted to comparing our results with them. Finally, various possible cues for the automatic detection of speech repairs--a key issue from the point of view of speech understanding--are explored: silent pauses, filled pauses, lengthenings, cut off words and discourse markers. As previously observed for English, none of them was found to be reliable by itself. More information, especially at the acoustic-prosodic level, is no doubt needed to reliably detect speech repairs.

Communication↗