Unknown Voice Research Articles

Cortical Responses to Mother's Voice in Comparison with Unfamiliar Voice in the First Trimester of Life: A fNIRS Study.

Introduction The use of functional near-infrared light spectroscopy (fNIRS) may be applied to study cortical responses in children and could offer insight into auditory and speech perception during the early stages of life. Some literature suggests that babies are already able to identify familiar voices at birth, and fNIRS is a non-invasive technique that can be used to study this population. Objective To characterize the cortical responses of infants during their first trimester of life to infant-directed speech using near-infrared light spectroscopy and to verify whether there is a difference in responses when infant-directed speech is performed by their mother compared with an unknown person. Methods Twenty-three children between 0 and 3 months, healthy, without risk indicators for hearing loss, and with results considered normal in the audiological evaluation were tested with near-infrared spectroscopy using infant-directed speech as an auditory stimulus produced by their own mother and by an unknown source. Results Bilateral cortical activation was observed. The responses were present in the temporal, frontal, and parietal regions. Regarding the familiarity aspect, no significant difference was observed for the mother's voice compared with an unknown voice. Conclusion Infant-directed speech has prosodic characteristics capable of activating several cortical regions in the infant's first trimester of life, especially the temporal region. The familiarity effect needs to be better defined for this type of stimulus during this period.

SpeakerNet for Cross-lingual Text-Independent Speaker Verification

Biometrics provide an alternative to passwords and pins for authentication. The emergence of machine learning algorithms provides an easy and economical solution to authentication problems. The phases of speaker verification protocol are training, enrollment of speakers and evaluation of unknown voice. In this paper, we addressed text independent speaker verification using Siamese convolutional network. Siamese networks are twin networks with shared weights. Feature space can be learnt easily by training these networks even if similar observations are placed in proximity. Extracted features from Siamese then can be classified using difference or correlation measures. We have implemented a customized scoring scheme that utilizes Siamese’ capability of applying distance measures with the convolutional learning. Experiments made on cross language audios of multi-lingual speakers confirm the capability of our architecture to handle gender, age and language independent speaker verification. Moreover, our designed Siamese network, SpeakerNet, provided better results than the existing speaker verification approaches by decreasing the equal error rate to 0.02.

Audio forensics behind the Iron Curtain: from raw sounds to expert testimony

ABSTRACT This essay investigates the construction of forensic audio expertise in the legal and security system of Communist Czechoslovakia and shows that the contested nature of speaker identification and sound-based objectivity contributed to the formulation of probabilistic claims in forensics. It explores the practices of the Department of Fonoscopy: a unique research laboratory of audio forensics that systematically examined the spectrographic, linguistic, and auditory means of sound analysis for the purpose of identifying unknown voices and environments in audio recordings. Bringing together the notions of “forensic cultures” and “sonic skills”, this article addresses the scientific, cultural, and political underpinning of the nascent field of audio expertise as well as the changing status of sound-based knowledge and forms of representation in forensics. In establishing fonoscopic expertise before the court and in the broader praxis of police investigation, the idea of vocal fingerprints and the use of sound visualisation technologies became instrumental. This essay pays special attention to the dynamics of the intricate process in which acoustic “raw material” (from anonymous calls, wiretapped phone lines, recorded conversations, or police interrogation rooms) was transformed into different kinds of legal and criminalistic evidence in the service of the totalitarian surveillance state.

Surviving Fifty Years With Shrapnel Within the Brain

AimsTraumatic brain injury during the war by shrapnel or bullet is lethal and life-threatening. The mortality rate from traumatic brain injury is more than 90%. Mrs. N (70) of Bangladesh is an exceptional and one of the luckiest women of the world. She is a war victim and has been living with shrapnel within her brain for the last 50 years.MethodsAccording to the patient attendant, half a century ago, during the Liberation War of Bangladesh, one shell accidentally exploded in their backyard. Unfortunately, a few pieces of shrapnel penetrated her arm, leg and right side of the head. She lost her consciousness and immediately treated at war hospital as per level best. She forgets details of her treatment and has no treatment records. Last 50 years, she has been surviving with movement difficulties, weaknesses of the upper and lower limbs and occasional convulsions. Her sufferings have intensified day by day. Last few years she has been experiencing headaches, dizziness, vomiting and forgetfulness. Two months back, she drank some insecticide mistakenly. She also suffering insomnia and she often cried out from deep sleep. She complains about hearing unknown voices. The voices were talking about her. She also started to suspect her family members. She claims that her family members were conspiring to kill her. She has history of convulsions three times in the last 50 years, which were generalized and tonic clonic in nature. She didn't take any medication for convulsion. As her suspicion, irrelevant talk and oddity in behavior worsened, her family members took her to me for treatment. Considering her past traumatic brain injury, an X-ray skull was advised. Surprisingly, it showed a metallic foreign body (shrapnel) within her skull. Furthermore, a CT scan of the brain was done, and it showed there was an irregular bony gap at right parietal area, a shrapnel at right frontal area with extensive encephalomalacic Porencephaly communicating with lateral ventricle. Her thyroid and liver function are normal. She is non-diabetic, non-hypertensive. She was advised to take risperidone 2 mg daily at night. With this low dose antipsychotic medication her psychosis controlled.ResultsSurviving with shrapnel within the brain followed by traumatic brain injury is very much rare. Shrapnel gradually damages the brain parenchyma and creates a large porencephaly.ConclusionTraumatic brain injury by bullet or shrapnel is always fatal. It is very much rare in medical to survive fifty years with shrapnel within the brain after shrapnel injury.

Revisiting the talker recognition advantage in bilingual infants

Recent work has shown that exposure to multiple languages affects nonlinguistic processing of speech during infancy. Specifically, Fecher and Johnson found that bilingual 9-month-olds outperformed their monolingual peers in a face–voice matching task in an unfamiliar language [Developmental Science (2019a), 22(4), e12778]. What factors were driving this effect? That is, was this finding truly reflective of a bilingual advantage specific to talker processing, or did the study demonstrate a general cognitive advantage in bilingual infants? Here, we revisited this question by testing bilingual and monolingual 9-month-olds (N = 48) on their ability to associate previously unknown voices with animated cartoon characters. In comparison with earlier work, where infants were presented with characters speaking an unfamiliar language (Spanish), the characters in this study spoke a language familiar to both groups of infants (English). Critically, we found that the monolingual and bilingual infants learned the face–voice pairings equally well when they were tested on the familiar language. We conclude that whereas bilingual infants are skilled at recognizing talkers regardless of the language spoken by the talkers, monolingual infants succeed at talker recognition in a familiar language only. These results begin to clarify the underlying nature of the talker recognition benefit previously reported for bilingual infants.

The processing of intimately familiar and unfamiliar voices: Specific neural responses of speaker recognition and identification.

Research has repeatedly shown that familiar and unfamiliar voices elicit different neural responses. But it has also been suggested that different neural correlates associate with the feeling of having heard a voice and knowing who the voice represents. The terminology used to designate these varying responses remains vague, creating a degree of confusion in the literature. Additionally, terms serving to designate tasks of voice discrimination, voice recognition, and speaker identification are often inconsistent creating further ambiguities. The present study used event-related potentials (ERPs) to clarify the difference between responses to 1) unknown voices, 2) trained-to-familiar voices as speech stimuli are repeatedly presented, and 3) intimately familiar voices. In an experiment, 13 participants listened to repeated utterances recorded from 12 speakers. Only one of the 12 voices was intimately familiar to a participant, whereas the remaining 11 voices were unfamiliar. The frequency of presentation of these 11 unfamiliar voices varied with only one being frequently presented (the trained-to-familiar voice). ERP analyses revealed different responses for intimately familiar and unfamiliar voices in two distinct time windows (P2 between 200–250 ms and a late positive component, LPC, between 450–850 ms post-onset) with late responses occurring only for intimately familiar voices. The LPC present sustained shifts, and short-time ERP components appear to reflect an early recognition stage. The trained voice equally elicited distinct responses, compared to rarely heard voices, but these occurred in a third time window (N250 between 300–350 ms post-onset). Overall, the timing of responses suggests that the processing of intimately familiar voices operates in two distinct steps of voice recognition, marked by a P2 on right centro-frontal sites, and speaker identification marked by an LPC component. The recognition of frequently heard voices entails an independent recognition process marked by a differential N250. Based on the present results and previous observations, it is proposed that there is a need to distinguish between processes of voice “recognition” and “identification”. The present study also specifies test conditions serving to reveal this distinction in neural responses, one of which bears on the length of speech stimuli given the late responses associated with voice identification.

Americanismos en Paraguay

In this paper we analyze El español en Paraguay by Manuel Alvar. We untangle from the Atlas those Americanisms that Spanish speakers do not know, and we compare them with the ones collected in both the Diccionario of the Royal Academy of Spanish Language, and the Diccionario de Americanismos from the same institution. In addition, we have compared those words with the ones collected in current Paraguayan dictionaries. The result is that unknown voices in Spain – or terms whose location is hard to define in lexicographical works – come to light.

A Novel automatic voice recognition system based on text-independent in a noisy environment

Automatic voice recognition system aims to limit fraudulent access to sensitive areas as labs. Our primary objective of this paper is to increase the accuracy of the voice recognition in noisy environment of the Microsoft Research (MSR) identity toolbox. The proposed system enabled the user to speak into the microphone then it will match unknown voice with other human voices existing in the database using a statistical model, in order to grant or deny access to the system. The voice recognition was done in two steps: training and testing. During the training a Universal Background Model as well as a Gaussian Mixtures Model: GMM-UBM models are calculated based on different sentences pronounced by the human voice (s) used to record the training data. Then the testing of voice signal in noisy environment calculated the Log-Likelihood Ratio of the GMM-UBM models in order to classify user's voice. However, before testing noise and de-noise methods were applied, we investigated different MFCC features of the voice to determine the best feature possible as well as noise filter algorithm that subsequently improved the performance of the automatic voice recognition system.

FanFutures

"FanFutures" is a project working with artificial (AI) automatic text generation in an exploration of fan fiction and speculation about possible futures. It is a collaboration between an artist, a computer scientist, and a social scientist — and, in an extended sense, also with the fan fiction community and an AI algorithm. Using a data-set of short sci-fi-esque stories written by anonymous members of the fan fiction community, we taught a natural language processing program to produce its own, entirely new short sci-fi-esque stories. The program takes imagined futures written by amateur writers, rather than institutionally-sanctioned voices, and dreams algorithmically through those voices to produce its own re-imaginings of possible futures. We then turned these stories into films, films that depict a computer’s dreams of the future drawn from a mass of unknown voices, using imagery selected by another algorithm. The outputs are feverish and confused, but, as with human dreams, we have embraced their incoherence, and allowed imagery and atmosphere to come to the fore. The project is inspired by our collective interest in understanding how humans make meaning with and through others — and how others make meaning with and through us. In a moment when our rapidly changing world, with its mass communication, new technologies and changing environment, seems almost intractable, we take a creative and playful approach to representing this complex intra-acting assemblage of human and more-than-human elements, and the possibilities in our world through sensory experience.

1. Higher Education and Alien Ecologies: Exploring the Dark Ontology of the University

The meaning of Anthropocene rests on the idea that there is a specific rationale behind the university and higher education, which is in itself progressive, educational, and redeeming. Institutions for, and students and teachers within, higher education, are more or less directly linked to the structures and rational of their political surroundings and social and cultural environments. However, just as there is depth to ecologies in the natural environment, and just as there is strangeness and even alien forms hidden within those environments, so it is with universities and higher education. They harbor hidden aspects, and these presences call from the dark, with their deeper and more unknown voices, even alien ones that seem to reach into universities not only from a future but also from a present that we cannot easily see or even understand. In this essay, we shed light on these more alien strands of higher education reality; what we term an “ontological excess” of universities. This darker ontology helps us argue that the foundations of the university go much deeper, and is societally more tangled, than the immediate pillars of the Anthropocene.

Journalist–source relations and the deliberative system: A network performance approach to investigating journalism’s contribution to facilitating public deliberation in a globalized world

Journalist–source relationships and interactions are interpreted in this study as crucial mechanisms for linking different arenas in a deliberative system. To unravel these source networks, 106 semi-standardized interviews with journalists as well as public relations (PR) professionals from government delegations and non-governmental organizations were conducted on-site three United Nations (UN) climate change conferences between 2010 and 2013, and an online survey was administered during the conference in 2015. The analysis shows that most journalists maintain close relationships with their home country delegation. However, journalists experienced in climate conference coverage also maintain more direct and informal relations to delegations from other countries and to non-governmental organizations while less experienced journalists exhibit loose and more formally mediated relationship to these actors. Moreover, journalists focusing on commentary rather than on event-related reporting have the most variegated and informal networks, thus opening the deliberative system to diverse perspectives and unknown voices more than others. Government delegations vary strongly in their tendency to approach journalists while environmental non-governmental organizations interact with journalists primarily to attract media attention in order to indirectly influence decision makers in national delegations.

Effect of mother's voice on neonatal respiratory activity and EEG delta amplitude.

While the influence of the mother's voice on neonatal heart‐rate response and its relevant activity on cerebral cortex and the autonomic nervous system (ANS) are well known, few studies have assessed its influence on respiratory activity. We investigated the relationship among the respiration rate, the delta wave amplitudes through electroencephalography, and the basal state of ANS through the respiratory variability index while 22 full‐term neonates hear their mother's voice and an unknown voice. It was found that when respiratory variability was large, a transient (<5 s) change in respiration rates was observed in response to an unknown voice, while a greater increase in the delta wave amplitude was observed in the frontal lobe than the parietal one in response to the mother's voice. Conversely, when respiratory variability was small, a sustained increase (>10 s) in respiration rates was observed in response to the mother's voice, while a greater increase in the delta wave amplitude was found in both the frontal and parietal lobes. These results suggest that the basal state of ANS influences the latency of increases in respiration rates. Furthermore, induced by the mother's voice, transient increases in respiration rates are reduced in association with frontal lobe activity, and sustained increases in respiration rates are promoted in association with frontal and parietal lobe activities.

Defective recognition and naming of famous people from voice in patients with unilateral temporal lobe tumours

Twenty-nine patients who underwent surgery for a temporal glioma, either in the left (16 patients) or right (13 patients) hemisphere, were administered standardized tests of unknown voice discrimination (UVD) and of famous voice recognition (VO-REC), which included tasks of familiarity evaluation, semantic identification and naming of famous voices. The UVD consisted of twenty stimuli, in which two audio files were consecutively presented; the subject was requested to judge whether the voices belonged to the same or different persons. In the VO-REC, patients were requested to recognize the voices of 40 very well known people; these voices were intermingled with the voices of 20 unknown people for a familiarity check, followed by identification and naming of persons recognized as familiar. We aimed at verifying the effect of laterality and intra-temporal site of lesion on familiarity assessment, false alarms, identification and naming of familiar people. As for the effect of lesion side, our results showed that patients with right temporal gliomas were significantly more impaired in voices discrimination and produced more false alarms than patients with a left glioma, who, in turn, were significantly more impaired in name retrieval than patients with a right temporal glioma. The high number of false alarms in patients with a right temporal glioma suggests that familiarity judgment was impaired. Regarding the neuroanatomical correlates of these different patterns of impairment, MRI data suggested that: (a) UVD disorders are due to lesions involving the whole right anterior temporal lobe and extending to lateral portions of the temporal and frontal lobes; (b) familiarity judgments (testified by an increased number of false alarms) are impaired in lesions restricted to the right anterior temporal lobe; (c) name retrieval deficits are found only in patients with left temporal lesions. UVD disorders were interpreted, at least in part, as due to an impairment of executive functions, resulting from a disconnection of the right temporal lobe from the frontal lobe control. A partly unexpected finding was that some patients with a right temporal tumour had a normal performance in famous voice recognition and identification, in spite of having severe voice discrimination disturbances. These unexpected results, in agreement with previous observation made in the visual (face) modality, are inconsistent with strictly hierarchical models of voice processing.

The “unknown voice” in Western history since Socrates

AbstractSocrates remains one of the most prominent paternal figures of Western dialogism and phonocentric paradigm; the man who stirred up the dialectic imaginations of his days. In Plato’s Socratic dialogues, his inner voice (daimonion) sounds as a last-instance statement to cast the light on the final solution of the conversation. In the context of antiquity and following cultural tradition, Socrates was the only hearer of warning signals from inside. The rest of the voices were urging (voices of the imaginable cursed souls, saints, angels, etc.). There was no need for a “personal” dictating voice when a divine dictation was already present. According to Charles Peirce’s classification, a “voice without evident source” or a voice from the head is a dicent indexical legisign.

Text Dependent Speaker Identification using Hidden Markchov Model and Mel Frequency Cepstrum Coefficient

Speaker identification is a biometric process. The objective of speaker identification is to extract, characterize and recognize the information about speaker identity. Speaker Recognition technology has recently been used in a vast number of commercial areas successfully such as in voice based biometrics; voice controlled appliances, security control for confidential information, remote access to computers and many more interesting areas. A speaker identification system has two phases which are the training phase and the testing phase. Feature extraction is the first step for each phase in speaker recognition. Many algorithms are used for feature extraction. In this work, the Mel Frequency Cepstrum Coefficient (MFCC) feature has been used for designing a text dependent speaker identification system. In the identification phase, the existing reference templates are compared with the unknown voice input. In this thesis, Hidden Markov Model (HMM) method is used as the training/recognition algorithm which makes the final decision about the specification of the speaker by comparing unknown features to all models in the database and selecting the best matching model. i, e. the highest scored model. The speaker who obtains the highest score is selected as the target speaker.

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.

Perception of pitch location within a speaker’s range: Fundamental frequency, voice quality and speaker sex

How are listeners able to identify whether the pitch of a brief isolated sample of an unknown voice is high or low in the overall pitch range of that speaker? Does the speaker's voice quality convey crucial information about pitch level? Results and statistical models of two experiments that provide answers to these questions are presented. First, listeners rated the pitch levels of vowels taken over the full pitch ranges of male and female speakers. The absolute f0 of the samples was by far the most important determinant of listeners' ratings, but with some effect of the sex of the speaker. Acoustic measures of voice quality had only a very small effect on these ratings. This result suggests that listeners have expectations about f0s for average speakers of each sex, and judge voice samples against such expectations. Second, listeners judged speaker sex for the same speech samples. Again, absolute f0 was the most important determinant of listeners' judgments, but now voice quality measures also played a role. Thus it seems that pitch level judgments depend on voice quality mostly indirectly, through its information about sex. Absolute f0 is the most important information for deciding both pitch level and speaker sex.

My Voice or Yours? An Electrophysiological Study

This study examined the neural processes underlying own voice discrimination using electrophysiological methods. Event-related potentials were recorded while healthy subjects (n=17) heard passively three oddball sequences composed of recordings of the French vowel/a/pronounced either by the participant her/himself or by two unknown persons. The results indicated that, although the mismatch negativity (MMN) displayed similar peak latency and amplitude in both conditions, the subsequent P3a clearly distinguished the two conditions since its amplitude was significantly smaller for own voice discrimination than for that of unknown voices. Moreover, the own voice discriminative response was associated with an early pre-MMN response. This early response involved a left inferior frontal component, the activity of which lasted throughout the time course of the discriminative response, which included both MMN and P3a.

Conclusion

Twenty-nine patients who underwent surgery for a temporal glioma, either in the left (16 patients) or right (13 patients) hemisphere, were administered standardized tests of unknown voice discrimination (UVD) and of famous voice recognition (VO-REC), which included tasks of familiarity evaluation, semantic identification and naming of famous voices. The UVD consisted of twenty stimuli, in which two audio files were consecutively presented; the subject was requested to judge whether the voices belonged to the same or different persons. In the VO-REC, patients were requested to recognize the voices of 40 very well known people; these voices were intermingled with the voices of 20 unknown people for a familiarity check, followed by identification and naming of persons recognized as familiar. We aimed at verifying the effect of laterality and intra-temporal site of lesion on familiarity assessment, false alarms, identification and naming of familiar people. As for the effect of lesion side, our results showed that patients with right temporal gliomas were significantly more impaired in voices discrimination and produced more false alarms than patients with a left glioma, who, in turn, were significantly more impaired in name retrieval than patients with a right temporal glioma. The high number of false alarms in patients with a right temporal glioma suggests that familiarity judgment was impaired. Regarding the neuroanatomical correlates of these different patterns of impairment, MRI data suggested that: (a) UVD disorders are due to lesions involving the whole right anterior temporal lobe and extending to lateral portions of the temporal and frontal lobes; (b) familiarity judgments (testified by an increased number of false alarms) are impaired in lesions restricted to the right anterior temporal lobe; (c) name retrieval deficits are found only in patients with left temporal lesions. UVD disorders were interpreted, at least in part, as due to an impairment of executive functions, resulting from a disconnection of the right temporal lobe from the frontal lobe control. A partly unexpected finding was that some patients with a right temporal tumour had a normal performance in famous voice recognition and identification, in spite of having severe voice discrimination disturbances. These unexpected results, in agreement with previous observation made in the visual (face) modality, are inconsistent with strictly hierarchical models of voice processing.

Teaching &amp; Learning Guide for: The Study of Variation from Two Perspectives

Teaching &amp; Learning Guide for: The Study of Variation from Two Perspectives