Speaker Characteristics Research Articles

아시아 상황의 영어 공통어 자질 연구: 리듬 특성

This paper investigated the rhythmic features of speakers of English as a lingua franca in Asian contexts. A speech corpus of 150 conversations between speakers of English in Asia with different L1 backgrounds was collected and %V, , VarcoV, and nPVI-V of each speaker were analyzed. It was found that L1 difference of the speakers and the speakers` daily use of English influenced %V, while the speakers` daily use of English influenced . The gender difference of the speakers also affected the rhythm of the utterances in VarcoV. A weak correlation between the two speakers` rhythm in each conversation was also found in %V and . No significant effects were found in nPVI-V. The results revealed that the speakers tended to accommodate the rhythm of their utterance to that of the interlocutors`. Further study on the speaking rate of the speakers is required to overcome some inconsistencies found in the results of the rhythmic metrics used in this study.

Older is wiser? It depends who you ask… and how you ask

ABSTRACTInteracting with older adults can trigger a number of stereotypes. Negative stereotypes tend to dominate, but a persistent positive perception of aging in the popular psyche is that age brings wisdom. Despite this common belief, previous research suggests that the relationship between age and wisdom is weak and inconsistent. We propose that this is due to opposing perceptions of aging – gains in life experience and socioemotional skills, counteracted by declines in cognitive skill – that also vary with characteristics of the perceiver and the context. We investigated these paradoxical perceptions of aging by asking younger (aged 20–40) and older (aged 60–80) adult judges to rate adult speakers (aged 30–89) on the basis of their responses to three narrative prompts. Several aspects of wisdom were judged, including those related to cognitive skills, and those related to socioemotional aspects of wisdom. We examined the impact on ratings of speaker and rater characteristics, as well as the interaction of these characteristics with the dimension being rated. Findings showed that the wisdom ratings were higher for speakers with better verbal fluency, and that the relationship between age and perceived wisdom was moderated by the speaker’s gender. Consistent with prior work, speaker age (relative to rater’s age) had a negative effect on cognitive aspects of wisdom, but a positive effect on socio-affective aspects of wisdom. Implications for understanding age-related stereotypes and conceptions of wisdom are discussed.

A Study of the Effect of Emotions and Software on Prosodic Features on Spoken Utterances in Urdu Language

Speech emotions have potential to provide valuable source of information which can lead us toward human perception and decision making process. This paper analyzes the variation and effect on prosodic features (Formant and Pitch) of female and male speakers in two different emotions (angry and neutral) and softwares (PRAAT and MATLAB) in Urdu language using two ways ANOVA testing. The objective of this paper is to determine the significant effect of emotions and softwares on prosodic features (Pitch and Formant) using recorded speech emotion of both male and female voices of same age group in Urdu language. Experimental results of two-way ANOVA testing considerably show that emotions have effect on pitch and formant both in male and female voice unlike software.

Effects of low-pass filtering on dialect and gender perception

In addition to linguistic (message-related) information, spoken language includes indexical information related to the speaker characteristics (e.g., gender, social status, and regional identity). This study is an extension of Jacewicz et al. (2015, JASA, 137, 2417–2418) which explored the nature of acoustic cues signaling indexical information. That study demonstrated that listeners were quite accurate in making decisions regarding the regional dialect and gender of a speaker when responding to short unprocessed phrases from 40 speakers (male and female) from two different dialects spoken in central Ohio and western North Carolina. However, when the signal was low-pass filtered at 400 Hz, sensitivity to both dropped significantly. The current study examined performance on the same phrases when the signal was low-pass filtered at progressively higher cutoff frequencies (500, 700, 900, and 1100 Hz). These stimuli were played to 20 listeners (10 males and 10 females) who identified the sex and dialect of the speaker. As expected, listener sensitivity improved with each wider low-pass filter condition. Performance in the 1100-Hz cutoff condition was very similar to that in the unprocessed condition. Discussion will focus on the nature of the acoustic cues utilized by listeners across the four low-pass filter conditions.

Speaker Frame Selection for Digital Speech Watermarking

Discriminative speaker features vary from one frame to another. Many researcher studied most of the discrimination features for speaker recognition by ignoring phonemic and linguistic information. However, other studies have proposed digital speech watermarking to improve the security of speaker recognition through channel transmission. Although digital speech watermark may be used with speaker information, it can affect speaker information which then can degrade the performance. This paper uses state-of-the-art in frame selection to propose a new way to preserve most of the discriminative features of speaker and to secure speech signals by applying digital speech watermarking technique. In this paper, linear predictive analysis was applied for each frame to extract formants, gain and residual errors. A frequency weighted function was used to quantify formants, and high order correlation as well as error gain is used for weighting the residual errors. The experimental results showed an overall (12 %) effectiveness in terms of performance, time and memory of frame selection technique for speaker recognition and digital speech watermarking.

Awareness of Racial Diversity in the Spanish‐Speaking World Among L2 Spanish Speakers

AbstractFor more than 30 years, foreign language instructors have considered approaches to teaching their students about Afro‐Latin cultures, yet little research appears to address students' actual awareness of these populations. Using a photo ratings survey, the present study explored the perceptions of postsecondary second language learners and their instructors concerning native Spanish speakers' physical appearance. The mean ratings assigned by the student participants to 24 photos were analyzed in conjunction with the following independent variables: first language, place of origin, race, gender, age, course level, contact with Spanish outside the classroom, having Spanish‐speaking friends or family, and travel to Spanish‐speaking countries. In addition, both students' and instructors' answers to open‐ended questions about native speaker characteristics were analyzed. Results indicated the importance for learners of having contact with Spanish outside the classroom, whether through exposure to media or encounters with Spanish speakers. Implications and suggestions are described for inclusive culture teaching.

밸런스드 아마추어 초소형 스피커의 진동 및 음향특성 연구

This paper describes the development process for vibration and acoustic characteristics of a balanced armature speaker. The design parameters were chosen in consideration of the influence of the bending stiffness of balanced armature which is the form of a cantilever structure in the speaker. For study of the performance of the speaker according to the design parameters, in the first step, we analyzed the characteristics of the velocity of the diaphragm to the electrical input. Next step, acoustic characteristics were analyzed by structural-acoustic coupled analysis. And the reliability of the analysis was verified by comparing the result of analysis with test results. Finally, we proposed a design method for implementing an enhanced balanced armature speakers through analysis method.

Analysis of Deaf Speakers’ Speech Signal for Understanding the Acoustic Characteristics by Territory Specific Utterances

An important concern with the deaf community is inability to hear partially or totally. This may affect the development of language during childhood, which limits their habitual existence. Consequently to facilitate such deaf speakers through certain assistive mechanism, an effort has been taken to understand the acoustic characteristics of deaf speakers by evaluating the territory specific utterances. Speech signals are acquired from 32 normal and 32 deaf speakers by uttering ten Indian native Tamil language words. The speech parameters like pitch, formants, signal-to-noise ratio, energy, intensity, jitter and shimmer are analyzed. From the results, it has been observed that the acoustic characteristics of deaf speakers differ significantly and their quantitative measure dominates the normal speakers for the words considered. The study also reveals that the informative part of speech in a normal and deaf speakers may be identified using the acoustic features. In addition, these attributes may be used for differential corrections of deaf speaker’s speech signal and facilitate listeners to understand the conveyed information.

Text-Independent Speaker Identification Using the Histogram Transform Model

In this paper, we propose a novel probabilistic method for the task of text-independent speaker identification (SI). In order to capture the dynamic information during SI, we design super-mel-frequency cepstral coefficients (MFCCs) features by cascading three neighboring MFCCs frames together. These super-MFCC vectors are utilized for probabilistic model training such that the speaker's characteristics can be sufficiently captured. The probability density function (PDF) of the aforementioned super-MFCCs features is estimated by the recently proposed histogram transform (HT) method. To recede the commonly occurred discontinuity problem in multivariate histograms computing, more training data are generated by the HT method. Using these generated data, a smooth PDF of the super-MFCCs vectors is obtained. Compared with the typical PDF estimation methods, such as Gaussian mixture model, promising improvements have been obtained by employing the HT-based model in SI.

“How Did That Make You Feel?” Influences of Gender and Parental Personality on Family Emotion Talk

Parent–child conversations about shared events may facilitate important aspects of psychosocial development through the family's references to emotions. Although past research has begun examining features of speakers and topics that predict emotion talk in these conversations, the potential contribution of personality traits has been overlooked. The current study investigated whether parental personality and participant gender were related to emotion talk during conversations between parents and their 5-year-old children. Results revealed discussion differences as a function of gender and discussion topic. Moreover, higher levels of trait Social Closeness in mothers corresponded to their richer emotional talk during discussions, particularly when conversing with daughters. Children's emotion talk regarding negative events was associated with their mothers' Positive Emotionality. No significant associations emerged for fathers' personality. The findings suggest that personality traits and gender may influence the very words that parents and children use, thus shaping the developmental environment in important ways.

Comparative Analysis of Vocal Characteristics in Speakers with Depression and High-Risk Suicide

Comparative Analysis of Vocal Characteristics in Speakers with Depression and High-Risk Suicide

High quality voice conversion using prosodic and high-resolution spectral features

Voice conversion methods have advanced rapidly over the last decade. Studies have shown that speaker characteristics are captured by spectral feature as well as various prosodic features. Most existing conversion methods focus on the spectral feature as it directly represents the timbre characteristics, while some conversion methods have focused only on the prosodic feature represented by the fundamental frequency. In this paper, a comprehensive framework using deep neural networks to convert both timbre and prosodic features is proposed. The timbre feature is represented by a high-resolution spectral feature. The prosodic features include F0, intensity and duration. It is well known that DNN is useful as a tool to model high-dimensional features. In this work, we show that DNN initialized by our proposed autoencoder pretraining yields good quality DNN conversion models. This pretraining is tailor-made for voice conversion and leverages on autoencoder to capture the generic spectral shape of source speech. Additionally, our framework uses segmental DNN models to capture the evolution of the prosodic features over time. To reconstruct the converted speech, the spectral feature produced by the DNN model is combined with the three prosodic features produced by the DNN segmental models. Our experimental results show that the application of both prosodic and high-resolution spectral features leads to quality converted speech as measured by objective evaluation and subjective listening tests.

Speaker-Adaptive Acoustic-Articulatory Inversion Using Cascaded Gaussian Mixture Regression

This paper addresses the adaptation of an acoustic-articulatory model of a reference speaker to the voice of another speaker, using a limited amount of audio-only data. In the context of pronunciation training, a virtual talking head displaying the internal speech articulators (e.g., the tongue) could be automatically animated by means of such a model using only the speaker’s voice. In this study, the articulatory-acoustic relationship of the reference speaker is modeled by a gaussian mixture model (GMM). To address the speaker adaptation problem, we propose a new framework called cascaded Gaussian mixture regression (C-GMR), and derive two implementations. The first one, referred to as Split-C-GMR, is a straightforward chaining of two distinct GMRs: one mapping the acoustic features of the source speaker into the acoustic space of the reference speaker, and the other estimating the articulatory trajectories with the reference model. In the second implementation, referred to as Integrated-C-GMR, the two mapping steps are tied together in a single probabilistic model. For this latter model, we present the full derivation of the exact EM training algorithm, that explicitly exploits the missing data methodology of machine learning. Other adaptation schemes based on maximum-a posteriori (MAP), maximum likelihood linear regression (MLLR) and direct cross-speaker acoustic-to-articulatory GMR are also investigated. Experiments conducted on two speakers for different amount of adaptation data show the interest of the proposed C-GMR techniques.

Computing scores of voice quality and speech intelligibility in tracheoesophageal speech for speech stimuli of varying lengths

In this paper, automatic assessment models are developed for two perceptual variables: speech intelligibility and voice quality. The models are developed and tested on a corpus of Dutch tracheoesophageal (TE) speakers. In this corpus, each speaker read a text passage of approximately 300 syllables and two speech therapists provided consensus scores for the two perceptual variables. Model accuracy and stability are investigated as a function of the amount of speech that is made available for speaker assessment (clinical setting). Five sets of automatically generated acoustic-phonetic speaker features are employed as model inputs. In Part I, models taking complete feature sets as inputs are compared to models taking only the features which are expected to have sufficient support in the speech available for assessment. In Part II, the impact of phonetic content and stimulus length on the computer-generated scores is investigated. Our general finding is that a text encompassing circa 100 syllables is long enough to achieve close to asymptotic accuracy.

Access the cluster tendency by visual methods for robust speech clustering

Identifying the cues for speech segments of speech data is an indispensable task in speaker clustering. The existing techniques perform the task of speech clustering without any prior knowledge of cluster tendency. Many techniques are investigated for finding a prior cluster tendency (CT). During the investigation, the visual access tendency (VAT) is recognized as a reasonable choice to find a cluster tendency. The speech clustering poses three important problems, which are as follows: modelling the speech data, cluster tendency, and effective speech clustering. Modelling is required for defining the shape of the speech segment based on the characteristics of speaker’s voice; hence it is useful for speech recognition. The GMM is a good choice for obtaining the precise model of speech data. Determining the number of speakers (or number of clusters) for the speech is known as cluster tendency. The quality of speech clustering depends on modelling and a prior clustering tendency. The classical algorithms [such as k-means, and minimum spanning tree (MST)-based-clustering] are merged with VAT for determining the effective clustering results along with a prior cluster tendency. We use linear subspace learning for representing the speech segments (or speech utterances) in a projected space of high-dimensional data. Various linear subspace learning techniques are used for improving the speech clustering results. The proposed approaches are hybrid approaches (i.e., k-means-CT, and MST–CT-based clustering), they use expensive steps. For this key reason, we propose another method, direct visualized clustering method, in which we derive the explicit speaker clustering results directly from VAT instead of using either k-means or MST-based clustering. We experimented the proposed methods on TSP speech datasets and done the comparative study for demonstrating the effectiveness of our work.

Characteristics of non-native speakers’ eyewitness reports on events: A comparison between a written report in the native language and an oral description in a second language

Characteristics of non-native speakers’ eyewitness reports on events: A comparison between a written report in the native language and an oral description in a second language

Electrophysiology of subject-verb agreement mediated by speakers' gender.

An important property of speech is that it explicitly conveys features of a speaker’s identity such as age or gender. This event-related potential (ERP) study examined the effects of social information provided by a speaker’s gender, i.e., the conceptual representation of gender, on subject–verb agreement. Despite numerous studies on agreement, little is known about syntactic computations generated by speaker characteristics extracted from the acoustic signal. Slovak is well suited to investigate this issue because it is a morphologically rich language in which agreement involves features for number, case, and gender. Grammaticality of a sentence can be evaluated by checking a speaker’s gender as conveyed by his/her voice. We examined how conceptual information about speaker gender, which is not syntactic but rather social and pragmatic in nature, is interpreted for the computation of agreement patterns. ERP responses to verbs disagreeing with the speaker’s gender (e.g., a sentence including a masculine verbal inflection spoken by a female person ‘the neighbors were upset because I ∗stoleMASC plums’) elicited a larger early posterior negativity compared to correct sentences. When the agreement was purely syntactic and did not depend on the speaker’s gender, a disagreement between a formally marked subject and the verb inflection (e.g., the womanFEM ∗stoleMASC plums) resulted in a larger P600 preceded by a larger anterior negativity compared to the control sentences. This result is in line with proposals according to which the recruitment of non-syntactic information such as the gender of the speaker results in N400-like effects, while formally marked syntactic features lead to structural integration as reflected in a LAN/P600 complex.

Sensitivity of automatic speaker identification to SVD digital audio watermarking

This paper proposes the utilization of SVD digital audio watermarking to increase the security of automatic speaker identification (ASI) systems and presents a study for the effect of watermarking on the ASI system performance. The SVD audio watermarking algorithm can be implemented on audio signals in time domain or in another appropriate transform domain and can be applied to the audio signal as a whole or on a segment-by-segment basis. The speaker recognition system works by generating a database of speaker's features using the MFCCs and polynomial shape coefficients extracted from each speaker after they are lexicographically ordered into 1-D signals. A matching process is performed for any new speaker to determine if he is belonging to the database or not, using a trained neural network. Experimental results show that the SVD audio watermarking doesn't degrade the ASI system performance severely. So, it can be used with ASI to increase security. Also, it was shown the segment by segment watermarking in the time domain achieves the highest detectability of the watermark. So, we can say that it is recommended to use SVD segment by segment audio watermarking with ASI systems implementing features extracted from the DCT or the DWT.

Voice conversion system using salient sub-bands and radial basis function

The objective of voice conversion is to replace the speaker-dependent characteristics of the source speaker so that it is perceptually similar to that of the target speaker. The speaker-dependent spectral parameters are characterized using single-scale interpolation techniques such as linear predictive coefficients, formant frequencies, mel cepstrum envelope and line spectral frequencies. These features provide a good approximation of the vocal tract, but produce artifacts at the frame boundaries which result in inaccurate parameter estimation and distortion in re-synthesis of the speech signal. This paper presents a novel approach of voice conversion based on multi-scale wavelet packet transform in the framework of radial basis neural network. The basic idea is to split the signal acoustic space into different salient frequency sub-bands, which are finely tuned to capture the speaker identity, conveyed by the speech signal. Characteristics of different wavelet filters are studied to determine the best filter for the proposed voice conversion system. A relative performance of the proposed algorithm is compared with the state-of-the-art wavelet-based voice morphing using various subjective and objective measures. The results reveal that the proposed algorithm performs better than the conventional wavelet-based voice morphing.

Speaker identification based on state space model

The dominant model for speaker recognition keeps Gaussian Mixture Model (GMM), though hidden Markov model (HMM) is successfully used in speech recognition. And then, few attentions have been paid on the correlation of features when modeling in the speaker recognition system, which may improve the final performance. Coincidentally, state-space model (SSM) is a model relying on the features correlation. A novel speaker identification method using a SSM based on the speaker features is proposed. The SSM, which has never been used in speaker identification before, presents the advantage of explicitly modeling spectral parameter trajectory dynamically. The preliminary system in our paper, directly regarding Mel frequency cepstrum coefficient (MFCC) as features, was set up to prove the correctness of SSM using in speaker recognition. The senior experiments, using supervectors, further explained the importance of the inner relationship of the speaker features, since considering SSM into GMM-UBM supervectors shows better accuracy, although just winning about 1 % in Arabic digital experiment.