Voice Patterns Research Articles

Identification Of Number Using Artificial Neural Network Backpropagation

This research proposed to design and implementation system of voice pattern recognition in the form of numbers with offline pronunciation. Artificial intelligent with backpropagation algorithm used on the simulation test. The test has been done to 100 voice files which got from 10 person voices for 10 different numbers. The words are consisting of number 0 to 9. The trial has been done with artificial neural network parameters such as tolerance value and the sum of a neuron. The best result is shown at tolerance value varied and a sum of the neuron is fixed. The percentage of this network training with optimal architecture and network parameter for each training data and new data are 82,2% and 53,3%. Therefore if tolerance value is fixed and a sum of neuron varied gave 82,2% for training data and 54,4% for new data

On the Use of Voice Signals for Studying Sclerosis Disease

Multiple sclerosis (MS) is a chronic demyelinating autoimmune disease affecting the central nervous system. One of its manifestations concerns impaired speech, also known as dysarthria. In many cases, a proper speech evaluation can play an important role in the diagnosis of MS. The identification of abnormal voice patterns can provide valid support for a physician in the diagnosing and monitoring of this neurological disease. In this paper, we present a method for vocal signal analysis in patients affected by MS. The goal is to identify the dysarthria in MS patients to perform an early diagnosis of the disease and to monitor its progress. The proposed method provides the acquisition and analysis of vocal signals, aiming to perform feature extraction and to identify relevant patterns useful to impaired speech associated with MS. This method integrates two well-known methodologies, acoustic analysis and vowel metric methodology, to better define pathological compared to healthy voices. As a result, this method provides patterns that could be useful indicators for physicians in identifying patients affected by MS. Moreover, the proposed procedure could be a valid support in early diagnosis as well as in monitoring treatment success, thus improving a patient’s life quality.

Automatic Baseball Video Tagging Based on Voice Pattern Prioritization and Recursive Model Localization

To enable us to select only the specific scenes that we want to watch in a baseball video and personalize its highlights sub-video, we require an Automatic Baseball Video Tagging system that can divide a baseball video into multiple sub-videos per at-bat scene automatically and append tag information relevant to at-bat scenes. Towards developing the system, the previous papers proposed several Tagging algorithms using ball-by-ball textual reports and voice recognition, and tried to refine models for baseball games. To improve its robustness, this paper proposes a novel Tagging method that utilizes multiple kinds of play-by-play comment patterns for voice recognition which represent the situation of at-bat scenes and take their “Priority” into account. In addition, to search for a voice-recognized play-by-play comment on the start/end of at-bat scenes, this paper proposes a novel modelling method called as “Local Modelling,” as well as Global Modelling used by the previous papers.

Acoustic Analysis of Normal Voice Patterns in Pakistani Adults

This study was designed to find out normative acoustic measures of adults in Pakistan. Age and gender differences in these parameters were also studied. The study variables include fundamental frequency, absolute jitter, relative jitter, relative average perturbation, 5-point period perturbation quotient perturbation quotient, jitter average absolute difference between consecutive periods, divided by the average period, shimmer local, shimmer decibel, three-point amplitude perturbation quotient, five-point amplitude perturbation quotient, average absolute difference of consecutive differences between the consecutive periods' amplitudes, and harmonics-to-noise ratio. Participants (N = 71; men = 34; women = 37) were recruited through two-stage sampling; convenience and purposive sampling. Participants were selected from students, staff, and faculty of the University of the Punjab, caretakers of the clients coming for clinical services, and researchers' friends, family, and acquaintances. Gender groups were further divided into three subgroups based on the age bands of 21-30, 31-40, and 41-50 years. After screening for health issues, the voice samples were recorded in a sound-proof acoustic laboratory by using a stand-held microphone and Praat software. Praat was also used for voice analyses. Means and standard deviations were conducted for the participants. Age and gender differences were investigated by using the two-way analysis of variance. Normative values were also measured for men and women between the age bracket of 21 and 50 years. Age differences were not significant for any variable. Gender differences were significant in fundamental frequency, absolute jitter, relative jitter, jitter period perturbation quotient, shimmer five-point amplitude perturbation quotient, and harmonics-to-noise ratio. The interaction between age and gender was significant in all the jitter measures. The interaction showed an increase in all these measures between the ages of 31 and 40 years and a decrease between the ages of 41 and 50 years in women, whereas men showed a reverse effect both in the age ranges of 31-40 and 41-50 years.

Agrammatism in Tagalog: voice and relativisation

ABSTRACTBackground: Virtually nothing is known about the ability of Tagalog speakers with agrammatic aphasia to cope with basic grammatical features of their language. Tagalog is unusual in exhibiting competing transitive patterns thanks to a system of voice that can make either of the verb’s arguments syntactically prominent – a prerequisite for undergoing syntactic operations such as relativisation.Aims: Our objective is to investigate the nature of the syntactic impairment associated with agrammatic aphasia in Tagalog, with special attention to voice patterns and relative clauses (RCs).Methods and Procedures: Five native Tagalog-speaking patients diagnosed with agrammatism took part in (1) elicited-production and (2) comprehension tasks to assess their ability with respect to voice patterns and (3) elicited-production, (4) imitation, and (5) comprehension tasks to assess their performance on RCs.Outcomes and Results: Although the participants did poorly on grammatical morphemes, the classic symptom of agrammatism, their impairment was somewhat less severe in patterns involving verb–agent–patient order, the patient-voice in declarative clauses, and the agent voice in RCs.Conclusions: Our findings reveal a departure from a general cross-linguistic tendency to favour agent-prominent constructions (e.g., the active voice in languages like English), as shown by the participants’ greater accuracy on patient voice declaratives than on their agent voice counterparts. Nonetheless, we found evidence of a subject preference in RCs, consistent with reports of a similar advantage in many other languages.

Using a Probabilistic Neural Network for lip-based biometric verification

In classical recognition techniques only raw features of objects are employed. Our approach allows use the composed features — so called Sim coefficients and landmarks which determine the area where biometric features should be searched. Biometric composed features are associated with appropriate similarity coefficients. Such approach brings significant advantages — recognition level of objects is higher compared to method based on the raw data. In this paper, a novel and effective lip-based biometric recognition approach with the Probabilistic Neural Network (PNN) is proposed. Lip based recognition has been less developed than the recognition of other human physical attributes such as the fingerprint, voice patterns, blood vessel patterns, or the face. For this reason, achieved results on this field are still improved and new recognition techniques are searched. Results achieved by PNN were improved by the Particle Swarm Optimization (PSO) technique.In the first step, lip area is restricted to a Region Of Interest (ROI) and in the second step, features extracted from ROI are specifically modeled by dedicated image processing algorithms. Extracted lip features are then an input data of neural network. All experiments were confirmed in the ten-fold cross validation fashion on three diverse datasets, Multi-PIE Face Dataset, PUT database and our own faces dataset. Obtained in researches result show that proposed approach achieves an average classification accuracy of 86.95%, 87.14%, and 87.26%, on these three datasets, respectively. Announced results were verified in the comparative studies and confirm the efficacy of the proposed lip based biometrics learned by PSO technique.

Vocal Change Patterns During a Teaching Day: Inter- and Intra-subject Variability

To describe the mean voice changes of 22 female teachers during a typical workday, examine the inter- and intra-subject variability, and establish a typology of different voice patterns during the workday. For each participant, fundamental frequency (F0), harmonics-to-noise ratio (HNR), jitter, and shimmer were measured on sustained vowels at the beginning and at the end of the workday, at three different times during the school year. The group mean pattern showed significant increases in F0 and HNR during the workday and significant decreases in jitter and shimmer. However, considerable inter- and intra-subject variability was observed. Based on the variation in the acoustic parameters during the workday, three different voice patterns were identified. The first is characterized by a greater F0 increase during the day, interpreted as a common, appropriate adaptation to vocal load. The second is characterized by a greater increase in HNR during the day and greater decreases in jitter and shimmer, interpreted as hyperfunctional voice production. The third is characterized by greater decreases in F0 and HNR and greater increases in jitter and shimmer, suggesting acute inflammation or muscle fatigue following the workday. The observed variety of vocal patterns during the workday emphasizes the need to study this phenomenon individually and target different types of behaviors to develop tailored prevention and treatment methods.

The intelligibility of speech in a harmonic masker varying in fundamental frequency contour, broadband temporal envelope, and spatial location

Differences in fundamental frequency (F0), modulations in the masker envelope, and differences in spatial location between a speech target and a masker can improve speech intelligibility in cocktail-party situations. These cues have been thoroughly investigated independently and associated with unmasking mechanisms: F0 segregation, temporal dip listening and spatial unmasking, respectively. Two experiments were conducted to examine whether F0 segregation interacts with spatial unmasking (experiment 1) or temporal modulations in the masker envelope (experiment 2) by measuring speech reception thresholds for a monotonized or an intonated voice against eight types of harmonic complex masker. In experiment 1, the masker varied in F0 contour (monotonized or intonated), mean F0 (0 or 3 semitones above that of the target) and spatial location (co-located or separated from the target). In experiment 2, the masker varied in F0 contour, mean F0 and broadband temporal envelope (stationary or 1-voice modulated). The benefits associated with spatial separation and F0 differences added up linearly in almost all conditions, whereas modulations in the masker envelope improved speech intelligibility only in the presence of intonated maskers. In addition, in both experiments F0 segregation benefited considerably from natural variations in the F0 pattern of the target voice, but was largely disrupted by those of the masker.

Tonal Organisation in Some F-MS Two-Voice Motets

The essay examines the tonal organisation of two-voice motets from Firenze, Biblioteca Medicea Laurenziana, ms. Pluteo 29.1. For each of them, the way in which the Tenor was extracted from chant, the melodic pattern of the upper voice, and the vertical relationships that subdivide the music continuum into discrete units are taken into account. The analysis reveals that (1) octoechos modality is a very weak structuring element of the musical flow, despite the modal-by-definition nature of their Tenores, and therefore (2) motet tonal organisation is not regulated by modal considerations, but by the force of the vertical interaction of the two voices. The comparison with conducti from the same Gothic repertoire that closes the essay, finally, shows that, paradoxically, the modal texture is weaker in motets than in conducti, even though these one have no connection to pre-existing liturgical melodies.

The speech signal segmentation algorithm using pitch synchronous analysis

Abstract Parameterization of the speech signal using the algorithms of analysis synchronized with the pitch frequency is discussed. Speech parameterization is performed by the average number of zero transitions function and the signal energy function. Parameterization results are used to segment the speech signal and to isolate the segments with stable spectral characteristics. Segmentation results can be used to generate a digital voice pattern of a person or be applied in the automatic speech recognition. Stages needed for continuous speech segmentation are described.

The bilingual voice: Vocal characteristics when speaking two languages across speech tasks

The present study compares vocal parameters in bilinguals’ use of two languages across different speech tasks, in order to examine whether language selection and task type will manifest different outcomes. Korean-English (KE) and Mandarin-English (ME) speakers performed three speech tasks – a reading passage, a monologue, and a picture description – in their two spoken languages. Fundamental frequency (F0), F0 variability, intensity, and speaking rate were measured. Results yielded an effect of language on average F0s, as both bilingual groups showed a significantly higher F0 in Mandarin or Korean compared to English. For the KE group, a higher F0 appeared in Korean compared to English across all tasks, while for the ME group, a difference in F0 between Mandarin and English was found for reading only. Both groups had the highest F0s in the reading task. There was a difference in F0 variability between two spoken languages only for the KE group monologue. Mean intensity of the ME group was overall higher than that of the KE group. Speaking rate was significantly slower in English than the native language for both groups, but the KE group demonstrated a greater difference in pace between their two spoken languages. Findings demonstrated that bilingual speakers produced notably different voice patterns contingent on language and speech task, indicating that inter- and intra-speaker variability in speakers’ vocal features can be attributed in part to language effects.

Rise and Demise of the Innovative Entrant

This story offers students the opportunity to practice systems thinking, including the analysis of behavior over time graphs, the application of systems archetypes, and the design of system interventions. Excerpt UVA-OM-1217 The Story Once upon a time, there was a small software company built around the idea of new innovative voice-recognition algorithm. This algorithm was much more efficient than those used in applications offered to the market by competitors—it was user-independent, allowed to shorten dramatically a setup period (period when a user “teaches” software his/her voice pattern), and could be developed into a language-independent version (i.e., interface working with exotic languages and dialects). The evident superiority of the algorithm over all existing product and prototypes prompted the company to develop its own set of applications and invest heavily in the marketing and sales network—they believed that given the proper marketing effort it would be easy to get people to switch to the new product and recruit new adepts of voice-recognition interface. The early years seemed to support the plan—sales were growing exponentially, every new application or release developed dominated the market, and the company further expanded its sales by educating customers on merits of using voice-recognition interface and tapping into new niches previously unexplored by competitors. . . .

Dysphonic Voice Pattern Analysis of Patients in Parkinson's Disease Using Minimum Interclass Probability Risk Feature Selection and Bagging Ensemble Learning Methods.

Analysis of quantified voice patterns is useful in the detection and assessment of dysphonia and related phonation disorders. In this paper, we first study the linear correlations between 22 voice parameters of fundamental frequency variability, amplitude variations, and nonlinear measures. The highly correlated vocal parameters are combined by using the linear discriminant analysis method. Based on the probability density functions estimated by the Parzen-window technique, we propose an interclass probability risk (ICPR) method to select the vocal parameters with small ICPR values as dominant features and compare with the modified Kullback-Leibler divergence (MKLD) feature selection approach. The experimental results show that the generalized logistic regression analysis (GLRA), support vector machine (SVM), and Bagging ensemble algorithm input with the ICPR features can provide better classification results than the same classifiers with the MKLD selected features. The SVM is much better at distinguishing normal vocal patterns with a specificity of 0.8542. Among the three classification methods, the Bagging ensemble algorithm with ICPR features can identify 90.77% vocal patterns, with the highest sensitivity of 0.9796 and largest area value of 0.9558 under the receiver operating characteristic curve. The classification results demonstrate the effectiveness of our feature selection and pattern analysis methods for dysphonic voice detection and measurement.

Real-time Simultaneous DKG and 2D DKG Using High-speed Digital Camera

For the evaluation of voice disorders, direct observation of vocal cord vibration is important. Among the various methods, laryngeal videostroboscopy (LVS) is widely used, but it was not a true image because it collects images from different cycles. In contrast, high-speed videoendoscopy and videokymography have much higher frame rates and can assess functional and mobility disorders. The purpose of the study is to describe real-time, simultaneous digital kymography (DKG), two-dimensional scanning (2D) DKG, and multi-frame (MF) LVS system using a high-speed digital camera, and identify the efficacy of this system in evaluating vibratory patterns of pathologic voice. The pattern of vocal fold vibration was evaluated in a vocally healthy subject and in subjects with vocal polyp, vocal nodules, vocal cord scar, and vocal cord paralysis. We used both quantitative (left-right phase symmetry, amplitude symmetry index) and qualitative (anterior-posterior phase symmetry) parameters for assessment of vocal fold vibration. Our system could record videos within seconds and required relatively little memory. The speed of replay in the DKG, 2D DKG, MF LVS, and high-speed videoendoscopy was controllable. The number of frame per cycle with MF LVS was almost the same as the fundamental frequency. Our system can provide images of various modalities simultaneously in real time and analyze morphological and functional vibratory patterns. It can be possible to provide a greater level of information for the diagnosis and treatment of vibratory disorders.

Differences of the Voice Parameters Between the Population of Different Hearing Tresholds: Findings by Using the Multi-Dimensional Voice Program.

Objectives To compare voice parameters in subjects with different hearing level.Methods The evaluation consisted of Multi-Dimensional Voice Program (MDVP) and electroglottography. Group 1 consisted of normal hearing subjects which is bilateral average hearing better than 25 decibels (dB) whereas group 2 consisted of patients who have bilateral average hearing between the 25 and 60 dB and group 3 consisted of patients who have bilateral average hearing between the 60 and 90 dB. The evaluations were performed on males and females separately.Results In female subjects, fundamental frequency (F0), absolute jitter, %jitter and soft phonation index (SPI) were significantly different between the group 1 and group 2. Also, we detected significant difference on maximum phonation time (MPT), fundamental frequency, absolute jitter and %jitter, and variable F0 (vF0) values between group 1 and group 3. Male subjects demonstrated significant difference between the group 1 and group 2 in MPT, absolute jitter, %jitter, vF0, and SPI parameters. Between the group 3 and group 1; differences in absolute jitter, %jitter, shimmer, %shimmer, vF0, and SPI were also significant.Conclusion This study concluded that even mild to moderate hearing losses may affect voice patterns in adults and also females and males react differently to hearing loss in some parameters.

A New Conceptual Approach for Voice Feminization: 12 Years of Experience.

Voice feminization has been developed with various surgical techniques to increase voice pitch. The purpose of this study was to evaluate 12 years of experience using a new endoscopical surgical technique, vocal fold shortening and retrodisplacement of the anterior commissure (VFSRAC) and retraining the voice pattern performed in female patients with androphonia or male-to-female transsexuals. Retrospective study. A retrospective study was performed on 362 patients who underwent VFSRAC (34 with androphonia, eight with androgenital syndrome, four with aplastic anemia treated with androgens, three with pseudohermaphroditism, and 313 male-to-female [MTF] transsexuals) and phonatory pattern rehabilitation between 2003 and 2014. Subjective and perceptual assessments, aerodynamic and acoustic assessments, and a videostroboscopic assessment were evaluated pre- and postoperatively in the female androphonia group and the MTF transsexual group. Mean preoperative fundamental frequency (F0 ) was 144.1 Hz, and mean postoperative F0 was 207.4 Hz, 6 months postoperatively. The mean increase of F0 in the female androphonia group was 53.0 Hz but was 73.6 Hz in the MTF transsexual group after the postoperative phonatory pattern retraining program. Voice femininity increased on the postoperative subjective assessment. Acoustic and aerodynamic assessments were within the normal range. Regularity of the vocal fold mucosal wave were maintained in the normal range. These findings suggest that patients could achieve a natural phonation pattern after surgery. Our new approach for voice feminization, VFSRAC, and retraining of the phonatory pattern was effective for patients who wanted a natural feminine voice. 4 Laryngoscope, 127:1102-1108, 2017.

Why Mobile Operators Introduced Data Plans: An Analysis of Voice and Data Usage Patterns

With the introduction of the data-oriented plan for LTE service, one may concerned with the background of the ISP’s policy in charging for LTE services. In this work we investigate the latest usage patterns of voice and data applications for customers over the current mobile network, via which we investigate why mobile operators introduced data-oriented plans. To be specific, we collected the real-field data for the volume of voice and data traffic from the LTE network before the dataoriented plans were introduced. From the collected data we compute the absolute volume as well as the proportion of voice and data applications. From these observations we infer mobile operators’ reasoning behind the decision to introduce dataoriented plans with unlimited voice calls over the mobile network.

The Patterns of Vowels in Monosyllabic Words of Uyghur Language

In this paper, on the basis of traditional phonetics, by using the methods of experimental phonetics and voice pattern theory, Analyzed and summarized the vowel pattern of the monosyllables in Uyghur language. The statistical analysis is carried upon the vowel formant frequency values in monosyllables, and discussed by using Joos method in more details. For the first time, with the actual experimental data proves the accordance of tongue location features of Uyghur vowel with the traditional knowledge from hearsay. The research results of this paper will have a high reference value for the study and application development of both Uyghur language and the other languages are belongs Altaic language family.

Estudo longitudinal de caracterização vocal em canto coral

RESUMO: Objetivo: verificar quais os benefícios do canto coral para o aprimoramento dos padrões de emissão vocal pré e pós atividade de voz cantada. Métodos: em três diferentes etapas, no período de um ano, foram coletados dados de sinais e sintomas vocais, registros de voz para análise perceptivo-auditiva e acústica, com os parâmetros de frequência fundamental, jitter, shimmer e ruído glótico, assim como a extensão vocal e o tempo máximo de fonação. A amostra foi composta por 23 cantores amadores de um coral universitário pertencente à Universidade Federal de Ciências da Saúde de Porto Alegre. A análise estatística consistiu de cruzar os dados em diferentes etapas de avaliação. Resultados: observou-se que a maioria dos cantores não apresentou padrões inadequados de uso vocal. Houve melhora da consciência de uso e cuidados vocais e de parâmetros de análise acústica. Conclusão: no período de um ano de mensuração, foi possível verificar que a prática de canto coral trouxe benefícios vocais aos sujeitos por meio do desenvolvimento e aprimoramento individual da emissão vocal.

Beyond the sensory: Findings from an in-depth analysis of the phenomenology of “auditory hallucinations” in schizophrenia

Objective: Research concerning the subjective sensory qualities of auditory hallucinations (AH) in people diagnosed with schizophrenia is scarce. Our goal was to investigate the “auditoriness” of AH and their overlap with symptoms grounded in alterations of thought rather than perception.Method: We undertook a detailed analysis of phenomenological interviews with 80 schizophrenia-spectrum voice-hearers.Results: We coded the dominant voice patterns of our subjects and found that only a minority (17.5%) reported a dominant pattern of AH which were experienced as literally auditory. Of dominant AH patterns, 11.3% were instead described as only quasi- or partially auditory, 28.8% as involving a combination of distinctly auditory and thought-like voices, and 15% as unambiguously thought-like. In addition, 5% reported exclusively simple, short-duration AH (e.g. hearing a single word), 12.5% the misperception of actual speech or sounds, and 10% predominantly multisensory voices. We also found substantial overlap between voices and symptoms traditionally considered abnormalities of thought rather than sensation.Conclusion: We believe these findings challenge common assumptions about AH in people diagnosed with schizophrenia, draw attention to potentially important but under-recognized characteristics of voices, and suggest a need for greater recognition of the heterogeneity of voices and the potential clinical as well as theoretical risks of conceptual over-simplification.