Analyzing training dependencies and posterior fusion in discriminant classification of apnoea patients based on sustained and connected speech

We present a novel approach using both sustained vowels and connected speech, to detect obstructive sleep apnea (OSA) cases within a homogeneous group of speakers. The proposed scheme is based on state-of-the-art GMM-based classifiers, and acknowledges specifically the way in which acoustic models are trained on standard databases, as well as the complexity of the resulting models and their adaptation to specific data. Our experimental database contains a suitable number of utterances and sustained speech from healthy (i.e control) and OSA Spanish speakers. Finally, a 25.1% relative reduction in classification error is achieved when fusing continuous and sustained speech classifiers. Index Terms: obstructive sleep apnea (OSA), gaussian mixture models (GMMs), background model (BM), classifier fusion.

Towards glottal source controllability in expressive speech synthesis

In order to obtain more human like sounding humanmachine interfaces we must first be able to give them expressive capabilities in the way of emotional and stylistic features so as to closely adequate them to the intended task. If we want to replicate those features it is not enough to merely replicate the prosodic information of fundamental frequency and speaking rhythm. The proposed additional layer is the modification of the glottal model, for which we make use of the GlottHMM parameters. This paper analyzes the viability of such an approach by verifying that the expressive nuances are captured by the aforementioned features, obtaining 95% recognition rates on styled speaking and 82% on emotional speech. Then we evaluate the effect of speaker bias and recording environment on the source modeling in order to quantify possible problems when analyzing multi-speaker databases. Finally we propose a speaking styles separation for Spanish based on prosodic features and check its perceptual significance.

Adapting a speech into sign language translation system to a new domain

This paper presents a methodology for adapting an advanced communication system for deaf people in a new domain. This methodology is a user-centered design approach consisting of four main steps: requirement analysis, parallel corpus generation, technology adaptation to the new domain, and finally, system evaluation. In this paper, the new considered domain has been the dialogues in a hotel reception. With this methodology, it was possible to develop the system in a few months, obtaining very good performance: good speech recognition and translation rates (around 90%) with small processing times.

Characterization of speech from amyotrophic lateral sclerosis by neuromorphic processing

Amyotrophic Lateral Sclerosis is a severe disease, which dramatically reduces the speech communication skills of patients as disease progresses. The present study is devoted to define accurate and objective estimates to characterize the loss of communication skills, to help clinicians and therapists in monitoring disease progression and in deciding on rehabilitation interventions. The methodology proposed is based on the perceptual (neuromorphic)definition of speech dinamics, concentrated in vowel sound in character and duration. We present the results from a longitudinal study carried out in an ALS patient during one year. Discussion addresses future actions.

A comparison of open-source segmentation architectures for dealing with imperfect data from the media in speech synthesis

Traditional Text-To-Speech (TTS) systems have been developed using especially-designed non-expressive scripted recordings. In order to develop a new generation of expressive TTS systems in the Simple4All project, real recordings from the media should be used for training new voices with a whole new range of speaking styles. However, for processing this more spontaneous material, the new systems must be able to deal with imperfect data (multi-speaker recordings, background and foreground music and noise), filtering out low-quality audio segments and creating mono-speaker clusters. In this paper we compare several architectures for combining speaker diarization and music and noise detection which improve the precision and overall quality of the segmentation.

Pilot study: Intranet communication among chronically ill hospitalized pediatric individuals

Hospitalization can be a very stressful experience, especially for children. With the use of technology, Intranet communication can be successful in obtaining interaction that these individuals lack to accomplish a positive adjustment to the hospital setting. The purpose of this exploratory, pilot project is to examine the use of networking chronically ill, hospitalized children with other hospitalized chronically ill children through Intranet communication.^ A target population of chronically ill hospitalized children, in at least Piaget's concrete operational stage, was asked to use the Intranet system to network with other chronically ill hospitalized children during their hospital stay, for one month or until discharge. The length of time of usage was recorded on a log sheet, and questionnaires were filled out at the end of the study.^ Statistical analysis was utilized to determine frequency of network usage, duration, demographics, and the impact on hospitalization. Results indicated that Intranet communication between chronically ill hospitalized children was utilized by the participants from 7-15 age groups; and had a positive impact on their hospitalization. ^

Cognitive and Visual Speech Contributions to Speech Perception in Challenging Listening Conditions

Speech perception routinely takes place in noisy or degraded listening environments, leading to ambiguity in the identity of the speech token. Here, I present one review paper and two experimental papers that highlight cognitive and visual speech contributions to the listening process, particularly in challenging listening environments. First, I survey the literature linking audiometric age-related hearing loss and cognitive decline and review the four proposed causal mechanisms underlying this link. I argue that future research in this area requires greater consideration of the functional overlap between hearing and cognition. I also present an alternative framework for understanding causal relationships between age-related declines in hearing and cognition, with emphasis on the interconnected nature of hearing and cognition and likely contributions from multiple causal mechanisms. I also provide a number of testable hypotheses to examine how impairments in one domain may affect the other. In my first experimental study, I examine the direct contribution of working memory (through a cognitive training manipulation) on speech in noise comprehension in older adults. My results challenge the efficacy of cognitive training more generally, and also provide support for the contribution of sentence context in reducing working memory load. My findings also challenge the ubiquitous use of the Reading Span test as a pure test of working memory. In a second experimental (fMRI) study, I examine the role of attention in audiovisual speech integration, particularly when the acoustic signal is degraded. I demonstrate that attentional processes support audiovisual speech integration in the middle and superior temporal gyri, as well as the fusiform gyrus. My results also suggest that the superior temporal sulcus is sensitive to intelligibility enhancement, regardless of how this benefit is obtained (i.e., whether it is obtained through visual speech information or speech clarity). In addition, I also demonstrate that both the cingulo-opercular network and motor speech areas are recruited in difficult listening conditions. Taken together, these findings augment our understanding of cognitive contributions to the listening process and demonstrate that memory, working memory, and executive control networks may flexibly be recruited in order to meet listening demands in challenging environments.

Malevolence, stigma, and social distance: Maximizing intergroup differences in HIV/AIDS discourse

This study examined the intergroup language used by young heterosexual Australians in conversations about HIV/AIDS and safe sex. Sixty male and 72 female heterosexuals participated in four-person facilitated conversations (same-sex or mixed-sex) about HIV/AIDS and safe sex, which were recorded and transcribed. We focused on extracts concerning strangers or malevolent individuals who appear to be group members, along with extracts involving foreign national groups. Discourse analysis showed that groups at lower levels of social distance were constructed mainly in terms of individual responsibility. At moderate social distance, stereotypes were more negative, but sub-typing was common, whereas at the highest levels, people were constructed entirely in intergroup terms. The findings of this study suggest that HN prevention programs should make reference to all salient outgroups, so as to neutralize communicative strategies that strengthen intergroup boundaries as a means of reducing perceived personal threat of HIV infection.

Homograph ambiguity resolution in front-end design for portuguese TTS systems

In this paper, a module for homograph disambiguation in Portuguese Text-to-Speech (TTS) is proposed. This module works with a part-of-speech (POS) parser, used to disambiguate homographs that belong to different parts-of-speech, and a semantic analyzer, used to disambiguate homographs which belong to the same part-of-speech. The proposed algorithms are meant to solve a significant part of homograph ambiguity in European Portuguese (EP) (106 homograph pairs so far). This system is ready to be integrated in a Letter-to-Sound (LTS) converter. The algorithms were trained and tested with different corpora. The obtained experimental results gave rise to 97.8% of accuracy rate. This methodology is also valid for Brazilian Portuguese (BP), since 95 homographs pairs are exactly the same as in EP. A comparison with a probabilistic approach was also done and results were discussed.

An automatic voice pleasantness classification system based on prosodic and acoustic patterns of voice preference

In the last few years the number of systems and devices that use voice based interaction has grown significantly. For a continued use of these systems the interface must be reliable and pleasant in order to provide an optimal user experience. However there are currently very few studies that try to evaluate how good is a voice when the application is a speech based interface. In this paper we present a new automatic voice pleasantness classification system based on prosodic and acoustic patterns of voice preference. Our study is based on a multi-language database composed by female voices. In the objective performance evaluation the system achieved a 7.3% error rate.

What your tot says about learning to talk

This leaflet was developed for parents and carers of babies aged between one and two years old who need help learning to talk.

La face et la figuration: une analyse interactionnelle tirée de l’instruction de base des Forces canadiennes

Le sujet de ce mémoire émane d’une volonté d’intégrer la recherche à ma réalité organisationnelle, celle des Forces canadiennes (FC). De cette aspiration est née l’idée d’étudier comment se dévoilent la face et la figuration lors de l’instruction de base. Partant d’une approche interactionniste, l’attention a été portée sur la face et la figuration d’un stagiaire au cours de trois interactions avec ses supérieurs. C’est plus précisément, la filature et l’analyse de conversation qui ont été mobilisées pour examiner cette problématique. La pertinence de la démarche est qu'elle met de l'avant une perspective de recherche peu présente dans la littérature, mais surtout, qu’elle s’insère dans un contexte organisationnel méconnu. L'entreprise vise d’une part, à nourrir la théorie en raffinant des concepts existants : la face et la figuration. D’autre part, elle cherche à éclairer la pratique en mettant en évidence des actes communicationnels difficilement observables dans le cadre habituel des activités militaires.

L’importance de la polyphonie organisationnelle dans un processus de changement stratégique

Ce mémoire de maîtrise s’applique à montrer comment la polyphonie narrative peut nous informer sur le vécu des employés en période de changement organisationnel. Pour être en mesure d’entendre ces multiples voix, nous avons fait l’étude des récits des employés et des dirigeants dans le cadre d’un tel processus. Nous voulions, à travers les récits des employés, illustrer empiriquement la manière dont ces récits constituent le changement organisationnel lui-même en mettant à jour les préoccupations qui alimentent le processus de changement et qui sont constitutives à la fois du changement et de l’entité organisationnels. Ces récits ont été analysés selon une perspective polyphonique, autrement dit, une approche mettant de l’avant l’existence de multiples voix en tensions, des voix composant ces récits. Pour procéder à l’analyse de ces récits, notre cadre théorique a mobilisé la notion de ventriloquie, tel que proposée et présentée dans les travaux de Cooren (2012, 2013). Ces préoccupations ont été analysées comme autant de voix animant et « auteurisant » (authoring) ces récits, fondant ainsi une polyphonie narrative que nous avons tenté de repérer et répertorier dans des entrevues. L’analyse de cette polyphonie par le repérage des voix, figures et tensions nous a permis de mettre à jour les préoccupations constitutives du changement lui-même. Dans ce mémoire, nous avons ainsi identifié puis analysé, les voix, les figures et les tensions qui illustrent le vécu de certains employés en période de changement organisationnel, comblant ainsi certaines lacunes repérées dans la littérature portant sur la polyphonie et le changement organisationnel. Comme nous le montrons, la littérature actuelle sur la polyphonie organisationnelle ne permet pas de saisir la manière dont les employés donnent voix à des préoccupations qui émergent du processus de changement et comment ces préoccupations interagissent et s’intègrent dans leurs récits organisationnels sur le changement. L’analyse de cette polyphonie par le repérage des voix, figures et tensions nous permet non seulement de mettre à jour les préoccupations constitutives du changement, mais aussi certains mécanismes communicationnels qui émergent de ce processus. En somme, l’étude de la polyphonie narrative nous permet de rendre compte de ce qui habite et préoccupe les employés en période de changement. Aussi, l’ensemble des préoccupations identifiées nous permet de brosser un portrait global du vécu des employés pendant cette période de changement tout en mettant à jour les tensions qui composent ce processus.

Le processus de l'écoute en consultation publique

Depuis une trentaine d’années, les citoyens des démocraties libérales boudent les isoloirs, souvent dépités par le manque d’impact de leur geste sur leur environnement (Sandel, 2012; Putnam, 2000). Plutôt que d’abandonner la sphère publique, plusieurs d’entre eux ont choisi de l’investir autrement. C’est ainsi que le monde occidental a connu un essor certain de la participation publique. Contrairement au vote, le citoyen est ici invité à prendre la parole, mais aussi à écouter les opinions des autres participants. Il se retrouve donc investi d’un rôle bien plus complexe sans être nécessairement doté des outils lui permettant d’accomplir la mission. Cette étude se penche sur le processus de l’écoute en consultation publique et vise à en déplier les mécanismes. Il s’agit d’observer comment s’exprime l’écoute en consultation publique, en dressant une typologie des marqueurs d’écoute et des sollicitations à l’écoute pour en étudier le fonctionnement. L’étude révèle qu’une écoute attentive peut se transformer en une écoute plus engagée quand une négociation mesurée de l’écoute prend place. Pour cela, chacun des acteurs doit être conscient de lui-même mais aussi de l’autre, du rôle de l’autre et du but final de la consultation. L’écoute engagée privilégie le partage d’autorité et auteurité nécessaire à la co-construction d’un discours entre tous les acteurs, une écoute qui se matérialise par le rapport que la commission remet aux preneurs de décision. La légitimité de ce rapport dépend du travail en amont entre les acteurs et, entre autres, de leur capacité à s’écouter, au delà de la dimension discursive. Le type d’écoute mis en œuvre contribue à inférer une légitimité à la consultation publique.