Utterance Duration Research Articles

Quality Measure Functions for Calibration of Speaker Recognition Systems in Various Duration Conditions

This paper investigates the effect of utterance duration to the calibration of a modern i-vector speaker recognition system with probabilistic linear discriminant analysis (PLDA) modeling. A calibration approach to deal with these effects using quality measure functions (QMFs) is proposed to include duration in the calibration transformation. Extensive experiments are performed in order to evaluate the robustness of the proposed calibration approach for unseen conditions in the training of calibration parameters. Using the latest NIST corpora for evaluation, results highlight the importance of considering the quality metrics like duration in calibrating the scores for automatic speaker recognition systems.

Analysing the Performance of Speaker Verification Task using Different Features

Speaker recognition is the identification of the person who is speaking by characteristics of their voices, also called “voice recognition”. The components of Speaker Recognition includes Speaker Identification(SI) and Speaker Verification(SV). Speaker identification is the task of determining an unknown speakers identity. If the speaker claims to be of a certain identity and the voice is to verify this claim, this is called Speaker Verification. It determines whether an unknown voice matches the known voice of a speaker whose identity is being claimed. This paper proposes Speaker Verification task. There are two phases in the Speaker Verification task namely, training and testing. In the training phase, different features such as Mel Frequency Cepstral Coefficient(MFCC), Linear Predictive Cepstral Coefficient(LPCC), Perceptual Linear Predictive(PLP) are extracted from the speech signal and is trained by Support Vector Machine to get the target speaker model. It is trained with both actual speaker and impostor utterances. In the testing phase, features are extracted from the test speech signal . The test features are extracted for different duration of time. The extracted feature vectors are given to the claimed speaker model and the decision is taken as authorised speaker or an impostor. The performance of a speaker verification task is analysed using different features with different utterance sizes. The result shows that the performance of a speaker verification task decreases when the duration of the speech utterances decreased.

Implication of silence in a Japanese psychotherapy context: a preliminary study using quantitative analysis of silence and utterance of a therapist and a client

We conducted a preliminary study to examine the effective index, which objectively indicates the development of the therapeutic process. It has been repeatedly reported that clients often insert silence in their speech when trying to focus on their feelings. This study quantitatively analysed the duration of pauses (i.e. silence) and utterances of a therapist and a client during Japanese psychotherapy sessions. Four, 50-minute sessions were analysed, including two low-evaluated sessions (low-evaluation cases) and two highly evaluated sessions (high-evaluation cases). Results indicated that the percentage and duration of clients’ pauses within high-evaluation cases were higher than those in low-evaluation cases. We discuss the implication of the use of silence during psychotherapy.

Looking to speak

This study investigates the different temporal orders that manifest in interactions involving a participant using an Augmentative Alternative Communications (AAC) device. Studies examining the use of AAC devices have regularly incorporated a particular understanding of temporality by using time as a measuring device to compare inter- and intra-individual action. In this paper we present an alternative perspective on time-in-interaction by showing how participants attend to time by examining their interactive behavior. Here, time is conceived in terms of how participants experience the duration and unfolding of a particular utterance. Through a close analysis of an interaction between a man with late-stage ALS and his wife, this paper shows how different orientations to time can underpin breakdowns of intersubjectivity. The analysis traces elements of this temporal disconnect to a variety of sources including the normative temporal expectations for the production of utterances through mouth-speech and the functioning of the device itself. The temporal misalignment leads to slippages in the participants’ orientation to the sequential relevance of utterances and utterance parts that leads to misunderstanding.

Development of speech motor control: Lip movement variability

This study examined variability of lip movements across repetitions of the same utterance as a function of age in Swedish speakers. The specific purpose was to extend earlier findings by examining variability in both phase and amplitude. Subjects were 50 typically developed native Swedish children and adults (28 females, 22 males, aged 5 to 31 yr). Lip movements were recorded during 15 to 20 repetitions of a short Swedish phrase using three-dimensional articulography. After correction for head movements, the kinematic records were expressed in a maxilla-based coordinate system. Movement onset and offset of the utterance were identified using kinematic landmarks. The Euclidean distance between receivers on the upper and lower lips was calculated and subjected to functional data analysis to assess both phase and amplitude variability. Results show a decrease in both indices as a function of age, with a greater reduction of amplitude variability. There was no difference between males and females for either index. The two indices were moderately correlated with each other, suggesting that they capture different aspects of speech production. Utterance duration also decreased with age, but variability was unrelated to duration. The standard deviation of utterance duration also decreased with age. The present results thus suggest that age related changes in speech motor control continue up until 30 years of age.

Embedded articulation: shifts in location influence speech production

During a conversation, the durations of spoken words decrease with repeated use, potentially reflecting increased accessibility of information encoded by the words. However, prior research has assessed this shortening only within single settings, despite the fact that conversations may continue across different locales. Importantly, movement through physical space has been shown to influence the accessibility of information encountered in prior events. This experiment assessed whether reductions in speech duration would be attenuated by changing settings. Participants completed a map route duplication task with a partner. Midway through the experiment, participants changed rooms, changed partners, changed both or changed neither. Over four trials, the durations of utterances for 10 key words were measured. As hypothesised, word duration showed a negative linear pattern among those not changing settings, but a cubic pattern for those changing rooms. Specifically, utterances showed a significant increase in duration immediately after shifting rooms. In contrast to the effect of the room manipulation, changing partners did not lessen reductions in word duration. These results highlight both the situated nature of dialogical speech and extend prior work on the cognitive outcomes associated with event boundaries.

ELEMENTY PROZODYCZNE W WOKALIZACJACH IMPLANTOWANYCH DZIECI GRECKOJĘZYCZNYCH PRZED WYKSZTAŁCENIEM MOWY

BackgroundThe speech prosody of young children who have CIs has not been the subject of detailed investigation. This case study is a longitudinal investigation of the vocalizations of two young children with CIs. The aims were to examine the suprasegmental characteristics of the prelinguistic utterances of these CI children and to develop a method for deriving phonoprosodic classifications using acoustical and auditory analyses.Material and MethodSpontaneous productions by two congenitally hearing-impaired Greek children fitted with a Nucleus-24 multichannel CI (ages 1: 10 and 2: 7; post-implant ages 0: 0 and 0: 6) were sampled for 6 months and, following transcription, classification of protophones was made. The analysis aimed to detect the frequency of occurrence of different protophone types in the infants’ utterances and to examine utterance duration and pitch contour via spectrography. Utterance characteristics were analyzed in relation to a) child’s age and b) post-implant age.ResultsAt post-implant age of 6 months, the younger and earlier implanted child showed more prelinguistic vocalizations and more complex structures than the older, later fitted child. The older, later CI recipient produced CV vocalizations at the beginning of implantation and gained a facility for frequent CV productions after 4 months of CI use.ConclusionsThe findings are in agreement with other studies which observe that later CI recipients produce CV combinations at earlier phases of the prelexical period until the necessary auditory input is provided by CIs. Protophone vocalizations serve as a comparative parameter of speech production level in pre-lingual speech.

A unified framework for domain independent online speaker indexing in eigen-voice space using an index tree of reference models

Speaker indexing referred in literature as speaker diarization is an important task in audio indexing and retrieval. Speaker indexing includes two important and usually separate stages, namely speaker segmentation and speaker clustering. Speaker indexing can be divided into online and offline categories. This paper mainly focuses on domain independent online speaker indexing. For this purpose, the proposed framework should be parameter free and no application specific parameters such as utterance duration or threshold settings are required. To reduce dependency on parameters, the traditional speaker segmentation is reformed to a voting based homogeneous speech segmentation, in which several approaches are applied in parallel to decide on the existence of a change point. In online indexing, data insufficiency is encountered at each time slice. In the proposed framework, a set of reference speaker models are used as side information to facilitate online tracking. To improve the indexing accuracy, adaptation approaches in eigen-voice decomposition space are proposed in this paper. To enhance the tracking performance from the computational cost point of view, an index structure of the reference models is proposed to speed up the search in the model space. The proposed framework is evaluated on the 2002 Rich Transcription Broadcast News and Conversational Telephone Speech Corpus (in Garofolo, NIST Rich Transcription, 2002) as well as a synthetic dataset. The indexing error of the proposed framework on telephone conversations, broadcast news and synthetic dataset are 7.51 %, 6.36 % and 9.34 %, respectively. Also, using the index tree structure approach, the tracking run time of the proposed framework is improved by 32 %.

Effects of practice schedules on speech motor learning

The aim of this study was to evaluate the effects of various practice schedules on learning a novel speech task. Forty healthy Cantonese speakers were asked to learn to produce a Cantonese phrase with two target utterance durations (2500 and 3500 milliseconds). They were randomly assigned to one of four learning conditions, each completing a different practice schedule, namely Blocked only, Random only, Blocked-then-Random, and Random-then-Blocked. Two retention tests (one immediate and one delayed) and a transfer test were administered. The four groups of participants showed different patterns of learning, but achieved comparable levels of performance at the end of the acquisition phase. However, participants in the Blocked only condition were less able to differentiate the two target durations than those in the Random only condition during retention. Furthermore, participants who received both blocked and random practice were less adversely affected by the secondary task during the transfer test than those who received either blocked or random practice alone. These findings suggest that mixed practice schedules are more effective than either blocked or random practice, especially in transferring the acquired speech motor skills to a cognitively demanding situation. The results have clinical implications regarding optimal practice schedules for treatment intervention.

観察者の存在が二者間の感情コミュニケーションに及ぼす影響

In everyday life, we communicate not only alone with another person but in front of others as well. Present study examined the effect of personal relationships between observer and dyad on facial behaviors during emotional communication. Pairs of friends viewed film clips that elicited positive affect and had a conversation in one of three conditions (a) observer absent, (b) stranger–observer, or (c) friend–observer. I video-recorded facial behavior of each participants and coded duration of smile, gaze and utterance. Participants filled out the questionnaire of emotion, public self-consciousness, internal other-consciousness, and satisfaction with conversation. Results showed that duration of utterance was longer in the friend–observer condition than the stranger–observer condition, but smiles and gazes did not differ among conditions. Observing by a friend seemed to facilitate emotional expressions on the intentional behavioral channel. Results of correlations between expressive behaviors and subjective indices suggest that friends made strategic use of nonverbal behavior in front of a stranger.

Temporal Relationship between Pause and Utterance Durations in Speech of Short Phrase

Temporal Relationship between Pause and Utterance Durations in Speech of Short Phrase

Prosodic Features of East Acehnese Dialect Based on Age Factor: An Experimental Phonetic Study

This study analyzed the prosodic features of East Acehnese dialect speech act, based on age factor of the speaker through Acoustic Phonetic Theory by using Praat Program. The target of this study was the directive speech in the form of command and offer. The findings of this study is that the speech tone of directive speech kapurono seudati ke jih based on basic tone, final tone, high tone, low tone, and temporal duration, uttered by the teenage informant is higher than that uttered by the adult informant. On the contrary , the speech tone of directive sentence tagun keumamah ke lon siat based on basic tone, final tone, low tone, and temporal duration, uttered by the adult informant is higher than that uttered by the teenage informant. Only in high tone, the speech tone uttered by the adult informant is lower than that uttered by the teenage informant. The duration of utterance uttered by teenage informant is longer than that uttered by the adult informant.

Longitudinal change in dysarthria associated with Friedreich ataxia: a potential clinical endpoint

CNS functions that show change across short periods of time are particularly useful clinical endpoints for Friedreich ataxia. This study determined whether there is measurable acoustical change in the dysarthria associated with Friedreich ataxia across yearly intervals. A total of 29 participants diagnosed with Friedreich ataxia were recorded across 4 years at yearly intervals. A repeated measures ANOVA was used to determine which acoustic measures differed across time, and pairwise t tests were used to assess the consistency of the change across the time intervals. The relationship between the identified measures with perceptual severity was assessed with stepwise regression. Significant longitudinal change was observed with four measures that relate to the utterance duration and spectral changes in utterances. The spectral measures consistently detected change across time intervals of two or more years. The four measures combined moderately predicted perceptual severity. Together, the results implicate longitudinal change in speaking rate and utterance duration. Changes in speech associated with Friedreich ataxia can be measured across intervals of 2 years and therefore show rich potential for monitoring disease progression and therapy outcomes.

Effect of an Observer's Presence on Facial Behavior During Dyadic Communication

In everyday life, people communicate not only with another person but also in front of other people. How do people behave during communication when observed by others? Effects of an observer (presence vs absence) and interpersonal relationship (friends vs strangers vs alone) on facial behavior were examined. Participants viewed film clips that elicited positive affect (film presentation) and discussed their impressions about the clips (conversation). Participants rated their subjective emotions and social motives. Durations of smiles, gazes, and utterances of each participant were coded. The presence of an observer did not affect facial behavior during the film presentation, but did affect gazes during conversation. Whereas the presence of an observer seemed to facilitate affiliation in pairs of strangers, communication between friends was exclusive and not affected by an observer.

Acoustic and perceptual measurements of prosody production on the profiling elements of prosodic systems in children by children with autism spectrum disorders

ABSTRACTProsody production atypicalities are a feature of autism spectrum disorders (ASDs), but behavioral measures of performance have failed to provide detail on the properties of these deficits. We used acoustic measures of prosody to compare children with ASDs to age-matched groups with learning disabilities and typically developing peers. Overall, the group with ASD had longer utterance durations on multiple subtests on a test of prosodic abilities, and both the ASD and learning disabilities groups had higher pitch ranges and pitch variance than the typically developing group on one subtest. Acoustic differences were present even when the prosody was used correctly. These findings represent differences in the fine details of the acoustic output beyond its functional interpretation inbothclinical groups.

Vowel undershoot in production of English tense and lax vowels by Mandarin and American speakers

Vowel undershoot (Lindblom, 1963), an effect where the articulatory gesture fails to reach the target due to the following contrary gesture (de Jong, 2004), is found to be particularly prominent (lowered F1 and shortened duration) in polysyllables (Lindblom, 1968; Moon and Lindblom, 1989). The current study investigates Mandarin and American speakers' production of English tense-lax vowels /i/-/I/ and /eI/-/ε/, and examines undershoot in monosyllabic and disyllabic words. Three Mandarin and three American speakers were recorded. Fifteen target vowels were embedded in a voiced stop-V-voiceless stop context. All disyllabic words were first-syllable stressed. Carrier sentences were of variable length to create a natural-speech-like context. F1, F2, vowel duration and utterance duration were measured. The results show that Americans did show a significant distinction between /i/-/I/ but tended to merge the formants of /eI/-/ε/ in disyllabic words, which was partly due to the coarticulation with the following syllable. They also demonstrated undershoot effect in /eI/-/ε/ but not much in /i/-/I /. Mandarin speakers, however, could not make a significant tense-lax distinction, and showed formant undershoot in disyllabic words, especially tense vowels. One possible account for this effect is the influence from L1. The issue of Mandarin vowel inventory is discussed.

Speech-associated labiomandibular movement in Mandarin-speaking children with quadriplegic cerebral palsy: A kinematic study

The purpose of this study was to investigate the speech-associated labiomandibular movement during articulation production in Mandarin-speaking children with spastic quadriplegic (SQ) cerebral palsy (CP). Twelve children with SQ CP (aged 7-11 years) and 12 age-matched healthy children as controls were enrolled for the study. All children underwent analysis of percentage of consonants correct (PCC) and kinematic analysis of speech tasks using the Vicon Motion 370 system. Kinematic parameters included utterance duration, displacement and velocity of the lip and jaw, coefficient of variation (CV) of lip utterance duration, and spatial and temporal coupling of labiomandibular movement of speech produced in mono-syllable (MS) and poly-syllable (PS) tasks. Children with CP showed lower temporal coupling (MS, p = 0.015; PS, p = 0.007), but not spatial coupling, of labiomandibular movement than healthy children. Children with CP had greater CVs (MS, p = 0.003; PS, p = 0.010) and the peak opening displacement and velocity of lower lip and jaw (p < 0.05) and lower PCC (p < 0.001) than healthy children. Children with SQ CP displayed labiomandibular coupling movement impairment, especially in the aspect of temporal coupling. These children also had high temporal oromotor variability and needed to make more effort to coordinate the labiomandibular movement for speech production.

Acoustic differences in the imitation of prosodic patterns in children with autism spectrum disorders

In research, it has been difficult to characterize the prosodic production differences that have been observed clinically in autism spectrum disorders (ASD). Moreover, the nature of these differences has been particularly hard to identify. This study examined one possible contributor to these perceived differences: motor planning. We examined the ability of children and adolescents with ASD to imitate prosodic patterns in comparison to a group with learning disabilities (LD) and a typically developing (TD) comparison group. Overall, we found that both the ASD and LD groups were significantly worse at perceiving and imitating prosodic patterns than the TD comparison group. Similar to previous studies using non-imitative speech, participants with ASD showed a significantly longer duration of utterances than the two comparison groups when attempting to imitate an intonation pattern. The implications of differences in duration of utterances are discussed. This study also highlights the importance of using clinical comparison groups in studies of language performance in individuals with ASD.

Tracking Talking: Dual Task Costs of Planning and Producing Speech for Young versus Older Adults

ABSTRACT A digital pursuit rotor was used to monitor speech planning and production costs by time-locking tracking performance to the auditory wave form produced as young and older adults were describing someone they admire. The speech sample and time-locked tracking record were segmented at utterance boundaries and multilevel modeling was used to determine how utterance-level predictors such as utterance duration or sentence grammatical complexity and person-level predictors such as speaker age or working memory capacity predicted tracking performance. Three models evaluated the costs of speech planning, the costs of speech production, and the costs of speech output monitoring. The results suggest that planning and producing propositionally dense utterances is more costly for older adults and that older adults experience increased costs as a result of having produced a long, informative, or rapid utterance.

Metrical regularity facilitates speech planning and production

AbstractProsodic structure is known to influence utterance production in numerous ways, but the influence of repetition of metrical pattern on utterance production has not been thoroughly investigated. It was hypothesized that metrical regularity would speed utterance production and reduce the occurrence of speech errors. Productions of sequences of four trisyllabic nonwords were compared between two conditions: a metrically regular condition with a repeating strong-weak-weak pattern, and a metrically irregular condition that lacked a repeating prominence pattern. Utterance durations were slower in the irregular condition, more hesitations occurred, and more sequencing errors were made. These findings are significant in that they are not accommodated by serial models of speech production. It is argued that the effects of metrical regularity are due to interference between words in an utterance plan, and that this interference arises from constraints on the dynamics of word form representations in the planning of speech.