No description available.

Ces travaux ont pour application la détection l'identification des signaux audio et particulièrement les signaux d'alarmes de voitures prioritaires. Dans un premier temps, nous proposons une méthode de détection des signaux d'alarme dans un environnement bruité, fondée sur des techniques d'analyse temps-fréquence des signaux. Cette méthode permet de détecter et d'identifier des signaux d'alarmes noyés dans du bruit, y compris pour des rapports signal à bruit négatifs. Puis nous proposons une quantification des signaux robuste aux bruits de transmission. Il s'agit de remplacer chaque niveau de bit d'un vecteur d'échantillons temporels ou fréquentiels par un mot binaire de même longueur fourni par un codeur correcteur d'erreur. Dans une première approche, chaque niveau de bits est quantifié indépendamment des autres selon le critère de minimisation de la distance de Hamming. Dans une seconde approche, pour réduire l'erreur de quantification à robustesse égale, les différents niveaux de bits sont quantifiés successivement selon un algorithme de type matching pursuit. Cette quantification donne aux signaux une forme spécifique permettant par la suite de les reconnaitre facilement parmi d'autres signaux. Nous proposons donc enfin deux méthodes de détection et d'identification des signaux fondées sur la quantification robuste, opérant dans le domaine temporel ou dans le domaine fréquentiel, par minimisation de la distance entre les signaux reçus restreints à leurs bits de poids fort et les signaux de référence. Ces méthodes permettent de détecter et d'identifier les signaux dans des environnements à rapport signal à bruit très faible et ceci grâce à la quantification. Par ailleurs, la première méthode, fondée sur la signature temps-fréquence, s'avère plus performante avec les signaux quantifiés. / This work targets the detection and identification of audio signals and in particular alarm signals from priority cars. First, we propose a method for detecting alarm signals in a noisy environment, based on time-frequency signal analysis. This method makes it possible to detect and identify alarm signals embedded in noise, even with negative signal-to-noise ratios. Then we propose a signal quantization robust against transmission noise. This involves replacing each bit level of a vector of time or frequency samples with a binary word of the same length provided by an error- correcting encoder. In a first approach, each bit level is quantized independently of the others according to the Hamming distance minimization criterion. In a second approach, to reduce the quantization error at equal robustness, the different bit levels are quantized successively by a matching pursuit algorithm. This quantization gives the signals a specific shape that allows them to be easily recognized among other signals. Finally, we propose two methods for detecting and identifying signals based on robust quantization, operating in the time domain or in the frequency domain, by minimizing the distance between the received signals restricted to their high-weight bits and the reference signals. These methods make it possible to detect and identify signals in environments with very low signal-to-noise ratios, thanks to quantization. In addition, the first method, based on the time-frequency signature, is more efficient with quantized signals.

A number of previous international studies have shown that many non-native Spanish speaking students lack sufficient oral communication skills. The aims of the study were twofold: 1) to investigate the possible effects of the Audio-Lingual method in the improvement of oral communication skills and 2) to test the effectiveness of two different teaching methods, i.e. the Audio-Lingual method and the Traditional method. A quantitative study was performed based on two different groups of students. One group worked with the Audio-Lingual method and the other group worked with the Traditional method. After a few lessons with the respective methods, the two groups were given a Fill-in the gap test that consisted of three parts: Greetings, Gender agreement and Number agreement. The test in Greetings and Number agreement showed no significant differences in results between the Audio-Lingual and the Traditional method. On the other hand, the test in Gender agreement showed that the Audio-Lingual method led to a significantly better result compared with the Traditional method. The study also demostrated that the students showed a greater interest in the Audio-Lingual method and this method helped the students to use their Spanish knowledge in other situations as well.

Ce travail porte sur la synthèse de la parole audio-visuelle. Dans la littérature disponible dans ce domaine, la plupart des approches traite le problème en le divisant en deux problèmes de synthèse. Le premier est la synthèse de la parole acoustique et l'autre étant la génération d'animation faciale correspondante. Mais, cela ne garantit pas une parfaite synchronisation et cohérence de la parole audio-visuelle. Pour pallier implicitement l'inconvénient ci-dessus, nous avons proposé une approche de synthèse de la parole acoustique-visuelle par la sélection naturelle des unités synchrones bimodales. La synthèse est basée sur le modèle de sélection d'unité classique. L'idée principale derrière cette technique de synthèse est de garder l'association naturelle entre la modalité acoustique et visuelle intacte. Nous décrivons la technique d'acquisition de corpus audio-visuelle et la préparation de la base de données pour notre système. Nous présentons une vue d'ensemble de notre système et nous détaillons les différents aspects de la sélection d'unités bimodales qui ont besoin d'être optimisées pour une bonne synthèse. L'objectif principal de ce travail est de synthétiser la dynamique de la parole plutôt qu'une tête parlante complète. Nous décrivons les caractéristiques visuelles cibles que nous avons conçues. Nous avons ensuite présenté un algorithme de pondération de la fonction cible. Cet algorithme que nous avons développé effectue une pondération de la fonction cible et l'élimination de fonctionnalités redondantes de manière itérative. Elle est basée sur la comparaison des classements de coûts cible et en se basant sur une distance calculée à partir des signaux de parole acoustiques et visuels dans le corpus. Enfin, nous présentons l'évaluation perceptive et subjective du système de synthèse final. Les résultats montrent que nous avons atteint l'objectif de synthétiser la dynamique de la parole raisonnablement bien / This work deals with audio-visual speech synthesis. In the vast literature available in this direction, many of the approaches deal with it by dividing it into two synthesis problems. One of it is acoustic speech synthesis and the other being the generation of corresponding facial animation. But, this does not guarantee a perfectly synchronous and coherent audio-visual speech. To overcome the above drawback implicitly, we proposed a different approach of acoustic-visual speech synthesis by the selection of naturally synchronous bimodal units. The synthesis is based on the classical unit selection paradigm. The main idea behind this synthesis technique is to keep the natural association between the acoustic and visual modality intact. We describe the audio-visual corpus acquisition technique and database preparation for our system. We present an overview of our system and detail the various aspects of bimodal unit selection that need to be optimized for good synthesis. The main focus of this work is to synthesize the speech dynamics well rather than a comprehensive talking head. We describe the visual target features that we designed. We subsequently present an algorithm for target feature weighting. This algorithm that we developed performs target feature weighting and redundant feature elimination iteratively. This is based on the comparison of target cost based ranking and a distance calculated based on the acoustic and visual speech signals of units in the corpus. Finally, we present the perceptual and subjective evaluation of the final synthesis system. The results show that we have achieved the goal of synthesizing the speech dynamics reasonably well

Dans cette thèse, nous traitons le problème de la séparation de sources audio multicanale par réseaux de neurones profonds (deep neural networks, DNNs). Notre approche se base sur le cadre classique de séparation par algorithme espérance-maximisation (EM) basé sur un modèle gaussien multicanal, dans lequel les sources sont caractérisées par leurs spectres de puissance à court terme et leurs matrices de covariance spatiales. Nous explorons et optimisons l'usage des DNNs pour estimer ces paramètres spectraux et spatiaux. À partir des paramètres estimés, nous calculons un filtre de Wiener multicanal variant dans le temps pour séparer chaque source. Nous étudions en détail l'impact de plusieurs choix de conception pour les DNNs spectraux et spatiaux. Nous considérons plusieurs fonctions de coût, représentations temps-fréquence, architectures, et tailles d'ensembles d'apprentissage. Ces fonctions de coût incluent en particulier une nouvelle fonction liée à la tâche pour les DNNs spectraux: le rapport signal-à-distorsion. Nous présentons aussi une formule d'estimation pondérée des paramètres spatiaux, qui généralise la formulation EM exacte. Sur une tâche de séparation de voix chantée, nos systèmes sont remarquablement proches de la méthode de l'état de l'art actuel et améliorent le rapport source-interférence de 2 dB. Sur une tâche de rehaussement de la parole, nos systèmes surpassent la formation de voies GEV-BAN de l'état de l'art de 14%, 7% et 1% relatifs en terme d'amélioration du taux d'erreur sur les mots sur des données à 6, 4 et 2 canaux respectivement / This thesis addresses the problem of multichannel audio source separation by exploiting deep neural networks (DNNs). We build upon the classical expectation-maximization (EM) based source separation framework employing a multichannel Gaussian model, in which the sources are characterized by their power spectral densities and their source spatial covariance matrices. We explore and optimize the use of DNNs for estimating these spectral and spatial parameters. Employing the estimated source parameters, we then derive a time-varying multichannel Wiener filter for the separation of each source. We extensively study the impact of various design choices for the spectral and spatial DNNs. We consider different cost functions, time-frequency representations, architectures, and training data sizes. Those cost functions notably include a newly proposed task-oriented signal-to-distortion ratio cost function for spectral DNNs. Furthermore, we present a weighted spatial parameter estimation formula, which generalizes the corresponding exact EM formulation. On a singing-voice separation task, our systems perform remarkably close to the current state-of-the-art method and provide up to 2 dB improvement of the source-to-interference ratio. On a speech enhancement task, our systems outperforms the state-of-the-art GEV-BAN beamformer by 14%, 7%, and 1% relative word error rate improvement on 6-channel, 4-channel, and 2-channel data, respectively

Submitted by Marlene Aparecida de Souza Cardozo (mcardozo@pucsp.br) on 2016-12-21T14:12:48Z No. of bitstreams: 6 Luiza Spínola Amaral.pdf: 1822997 bytes, checksum: c152b665d4104e27b7351b0243a1b8d8 (MD5) John Cage- 4'33'' for piano (1952).mp4: 24651231 bytes, checksum: fb65ab6cf2d80d1a2fca2b2a119c3b44 (MD5) 001.mp3: 14885300 bytes, checksum: ee5bdaa5462f6c619d99586e20b227a3 (MD5) 002.mp3: 10558049 bytes, checksum: a3ec7085b41e649147f9cb9a9923bb87 (MD5) 004.mp3: 25652929 bytes, checksum: b5208b6be64e183a617a88c0d17db616 (MD5) 005.mp3: 17304797 bytes, checksum: a514538aa17de813c7968a9ebb4f3ce2 (MD5) / Made available in DSpace on 2016-12-21T14:12:48Z (GMT). No. of bitstreams: 6 Luiza Spínola Amaral.pdf: 1822997 bytes, checksum: c152b665d4104e27b7351b0243a1b8d8 (MD5) John Cage- 4'33'' for piano (1952).mp4: 24651231 bytes, checksum: fb65ab6cf2d80d1a2fca2b2a119c3b44 (MD5) 001.mp3: 14885300 bytes, checksum: ee5bdaa5462f6c619d99586e20b227a3 (MD5) 002.mp3: 10558049 bytes, checksum: a3ec7085b41e649147f9cb9a9923bb87 (MD5) 004.mp3: 25652929 bytes, checksum: b5208b6be64e183a617a88c0d17db616 (MD5) 005.mp3: 17304797 bytes, checksum: a514538aa17de813c7968a9ebb4f3ce2 (MD5) Previous issue date: 2016-12-02 / Conselho Nacional de Desenvolvimento Científico e Tecnológico / The purpose of this research is to make a critical analysis about the literary and media work of the Germany radio man and jazz specialist, Joachim-Ernst Berendt. The focus is on the second stage of his career, no long dedicated to music, but to sounds and to hear, with the proposal to elaborate the concept of "audio-image". For this, he proposes a corpus that is composed by the radio masterpiece and then book "Nada Brahma – Die Welt ist Klang" (called "Nada Brahma: o mundo e o universo da consciência" in the Brazilian edition). With the justification that the visual images do not seduces the glances anymore due to the high speed of dissemination in contemporary cultural environments, this thesis investigates another approach about the image, able to contribute to a theory of audio images, which looks very necessary due to sound obscuring in the profusion of audio-visual media. In this sense, in addition to present the ideas of the German author in the domain of the communication's research in Brazil, this thesis defines the hearing experience as a resistance strength against the hegemony of visual images. Thus, it approaches the thought of Berendt to the Dietmar Kamper one, both in dialog with the thought of Norval Baitello Jr. in his theory of the image and culture. As his own reference, Baitello Jr. presents Berendt, extending the dialog with other thinkers as Harry Pross, which includes the body in its sensorial perception as primary media in its media classification, and Kamper himself, when announces the return to the hearing as a way of combating the "suffering" of the eyes. Thereby, the thesis proposes a concept of sound image close to a science called, by Hans Belting, "Anthropology of Image", where the sound appears as a vehicle of culture. For an understanding of the sounds and the media, the research uses the work of musicologists and sound engineers R. Murray Schafer and Michel Chion, which do not reduce the understanding of sound to the field of tradition of music. The theoretical framework also includes the authors cited by Berendt and his commentators, like the philosopher Peter Sloterdijk, the historians Andrew Right Hurley and Uta G. Poiger, in addition to the maestro and musicologist Julio Medaglia. We propose therefore a reflection about the restoration of relations and links broken in the light of the new culture of media / O objetivo desta pesquisa é fazer uma análise crítica acerca da obra literária e midiática do radialista alemão, especialista em jazz, Joachim-Ernst Berendt, com foco na segunda fase da sua carreira, não mais dedicada à música, mas aos sons e ao ouvir, com a proposta de elaborar o conceito de ‘áudio-imagem’. Para tanto, propõe um corpus composto pela peça radiofônica e depois livro, Nada Brahma – Die Welt ist Klang (denominado na edição brasileira, Nada Brahma: o mundo e o universo da consciência). Com a justificativa de que as imagens visuais já não mais seduzem os olhares devido à alta velocidade de disseminação nos ambientes culturais contemporâneos, essa tese investiga outra abordagem acerca da imagem, capaz de contribuir com uma teoria das imagens auditivas, que parece muito necessária frente o ofuscamento sonoro na profusão das mídias audiovisuais. Nesse sentido, além de apresentar as ideias do autor alemão no âmbito das pesquisas de comunicação no Brasil, esta tese define a experiência auditiva como força de resistência à hegemonia das imagens visuais. Aproxima, assim, o pensamento de Berendt ao de Dietmar Kamper, em diálogo com o pensamento de Norval Baitello Jr. na sua teoria da imagem e da cultura. Como própria referência, Baitello Jr. apresenta Berendt, ampliando o diálogo com outros pensadores como Harry Pross, que na sua classificação da mídia inclui o corpo em sua percepção sensorial como mídia primária e o próprio Kamper, quando anuncia o retorno à audição como forma de combate ao “padecimento” dos olhos. Assim, a tese propõe um conceito de imagem sonora próximo de uma ciência denominada, por Hans Belting, “Antropologia da Imagem”, onde o som aparece como veículo da cultura. Para uma compreensão acerca dos sons e da mídia, a pesquisa se vale dos trabalhos dos musicólogos e engenheiros de som R. Murray Schafer e Michel Chion, que não reduzem o entendimento de som ao campo da tradição da música. O referencial teórico inclui ainda os autores citados por Berendt e seus comentadores, como o filósofo Peter Sloterdijk, os historiadores Andrew Right Hurley e Uta G. Poiger, além do maestro e musicólogo Júlio Medaglia. Propomos, assim, uma reflexão acerca do restabelecimento de nexos e vínculos rompidos frente à nova cultura das mídias

Submitted by Filipe dos Santos (fsantos@pucsp.br) on 2017-01-11T15:45:44Z No. of bitstreams: 1 Lucila Lerro Rupp.pdf: 2480728 bytes, checksum: 3f4db3906ccf5192e942f8a5d2dd513d (MD5) / Made available in DSpace on 2017-01-11T15:45:44Z (GMT). No. of bitstreams: 1 Lucila Lerro Rupp.pdf: 2480728 bytes, checksum: 3f4db3906ccf5192e942f8a5d2dd513d (MD5) Previous issue date: 2016-12-16 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / The main aim of this Masters dissertation is to understand how educational videos can be attached to the curriculum and what elements it should contain. The work also tries to find some of these elements in audio-visual scripts. The dissertation has a qualitative approach and uses the following techniques: documental research, script analysis and four experiments conducted in public schools located in the city of São Paulo. The research shows that students acknowledge the videos as curriculum mediators; are able to identify specific characteristics of the media; and that these productions, when properly used by the teacher, are well received by pupils, enriching the reflection on the curriculum content / O objetivo geral deste trabalho é compreender como o vídeo educacional pode se integrar ao currículo e quais devem ser os elementos desse recurso. O objetivo específico é descobrir se é possível delinear algumas dessas características no roteiro do audiovisual. O estudo se orienta pelos princípios da abordagem qualitativa, e sua metodologia envolve pesquisa documental, análise de roteiros de vídeos educacionais e experiências em quatro escolas públicas municipais de São Paulo. A pesquisa mostra que os alunos reconhecem no vídeo um mediador curricular; identificam características específicas do meio audiovisual; e que as produções, quando apropriadas e contextualizadas pelo professor, são bem-vindas na sala de aula, enriquecendo as possibilidades de reflexão e de entendimento dos conteúdos curriculares

Made available in DSpace on 2016-04-26T18:18:03Z (GMT). No. of bitstreams: 1 Johan Cavalcanti van Haandel.pdf: 1998712 bytes, checksum: 208cad61d2ab422d9484c67e1d70027c (MD5) Previous issue date: 2009-05-22 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / In the last years the transmission of audio is being transformed for the new technologies of the digital support, which had generated new processes and products, among them is the sonorous webcasting. This work investigates the creation of content of the four existing formats of sonorous webcasting: web radio, playlist, on demand audio and the audio portal, which comprises the corpus of the research. As methodological strategy, firstly, examples of the sonorous webcasting formats and of others transmission processes for the digital support had been observed, to distinguish them; the second step was to observe the assembly of audio of the sonorous webcasting formats and the third step was to investigate its graphical interfaces. With basis in the methodological strategy, the research presents itself in three stages: in the first one, it was outlined a mapping of the existing digital transmissions of audio and a classification of sonorous webcasting formats, in which the main authors of reference are Trigo-De-Souza and Kischinhevsky, among others; in the second stage, the construction of the transmission of sonorous webcasting through audio assembly is investigated with data based on Ferraretto, Cyro César, Barbosa Filho and McLeish, among others; and in third stage, it is investigated the construction of the transmission of sonorous webcasting through the graphical interface, where the hypermedia interfaces study referring data are based on Beiguelman, Manovich and Johnson. It is concluded that sonorous webcasting consists of a transmission process of audio that does not have to be confounded with radio and it sets up a new type of content reading, mediated by texts and images, implying, therefore, in studies of audio assembly and graphical interface / Nos últimos anos a transmissão de áudio está sendo transformada pelas novas tecnologias do suporte digital, as quais geraram novos processos e produtos, entre eles o webcasting sonoro. A presente pesquisa investiga a criação de conteúdo dos quatro formatos existentes do webcasting sonoro: a web rádio, a playlist, o áudio on demand e o portal de áudio, os quais compõem o corpus da pesquisa. Como estratégia metodológica, primeiro foram observados exemplos dos formatos do webcasting sonoro e de outros processos de transmissão pelo suporte digital, para particularizá-los; o segundo passo foi observar a montagem de áudio dos formatos do webcasting sonoro e o terceiro passo foi investigar as suas interfaces gráficas. Baseada na estratégia metodológica, a pesquisa se apresenta em três etapas: na primeira, há um mapeamento das transmissões digitais de áudio existentes e uma classificação dos formatos do webcasting sonoro, em que os principais autores de referência são Trigo-De-Souza e Kischinhevsky, entre outros; na segunda, é investigada a construção da transmissão do webcasting sonoro através da montagem de áudio, com dados baseados em Ferraretto, Cyro César, Barbosa Filho e McLeish, entre outros; e na terceira, é investigada a construção da transmissão do webcasting sonoro através da interface gráfica, em que os dados referentes ao estudo de interfaces da hipermídia são baseados em Beiguelman, Manovich e Johnson. Conclui-se que o webcasting sonoro consiste em um processo de transmissão de áudio que não deve ser confundido com rádio e que institui um novo tipo de leitura de conteúdo, mediado por textos e imagens, implicando, por isso, em estudos de montagem sonora e de interface gráfica

Tese de mestrado. Multimédia (Música Interactiva e Design de Som). Faculdade de Engenharia. Universidade do Porto. 2012

Les logiciels de patching audio traditionnels, tels que Max ou Pure Data, sont des environnements qui permettent de concevoir et d’exécuter des traitements sonores en temps réel. Ces logiciels sont mono-utilisateurs, or, dans bien des cas, les utilisateurs ont besoin de travailler en étroite collaboration à l’élaboration ou à l’exécution d’un même traitement. C’est notamment le cas dans un contexte pédagogique ainsi que pour la création musicale collective. Des solutions existent, mais ne conviennent pas forcément à tous les usages. Aussi avons-nous cherché à nous confronter de manière concrète à cette problématique en développant une nouvelle solution de patching audio collaborative, baptisée Kiwi, qui permet l’élaboration d’un même traitement sonore à plusieurs mains de manière distribuée. À travers une étude critique des solutions logicielles existantes nous donnons des clefs de compréhension pour appréhender la conception d’un système multi-utilisateur de ce type. Nous énonçons les principaux verrous que nous avons eu à lever pour rendre cette pratique viable et présentons la solution logicielle. Nous exposons les possibilités offertes par l’application et les choix de mise en œuvre techniques et ergonomiques que nous avons faits pour permettre à plusieurs personnes de coordonner leurs activités au sein d’un espace de travail mis en commun. Nous revenons ensuite sur différents cas d’utilisation de ce collecticiel dans un contexte pédagogique et de création musicale afin d’évaluer la solution proposée. Nous exposons enfin les développements plus récents et ouvrons sur les perspectives futures que cette application nous permet d’envisager. / Traditional audio patching software, such as Max or Pure Data, are environments that allow you to design and execute sound processing in real time. These programs are single-user, but, in many cases, users need to work together and in a tight way to create and play the same sound processing. This is particularly the case in a pedagogical context and for collective musical creation. Solutions exist, but are not necessarily suitable for all uses. We have tried to confront this problem in a concrete way by developing a new collaborative audio patching solution, named Kiwi, which allows the design of a sound processing with several hands in a distributed manner. Through a critical study of the existing software solutions we give keys of comprehension to apprehend the design of a multi-user system of this type. We present the main barriers that we had to lift to make this practice viable and present the software solution. We show the possibilities offered by the application and the technical and ergonomic implementation choices that we have made to allow several people to coordinate their activities within a shared workspace. Then, we study several uses of this groupware in pedagogical and musical creation contexts in order to evaluate the proposed solution. Finally, we present the recent developments and open up new perspectives for the application.