• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 170
  • 40
  • 33
  • 30
  • 14
  • 10
  • 9
  • 8
  • 4
  • 4
  • 4
  • 3
  • 3
  • 2
  • 2
  • Tagged with
  • 391
  • 104
  • 101
  • 86
  • 80
  • 47
  • 39
  • 33
  • 32
  • 31
  • 30
  • 30
  • 28
  • 28
  • 27
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
131

Reconhecimento automático do locutor com redes neurais pulsadas. / Automatic speaker recognition using pulse coupled neural networks.

Timoszczuk, Antonio Pedro 22 March 2004 (has links)
As Redes Neurais Pulsadas são objeto de intensa pesquisa na atualidade. Neste trabalho é avaliado o potencial de aplicação deste paradigma neural, na tarefa de reconhecimento automático do locutor. Após uma revisão dos tópicos considerados importantes para o entendimento do reconhecimento automático do locutor e das redes neurais artificiais, é realizada a implementação e testes do modelo de neurônio com resposta por impulsos. A partir deste modelo é proposta uma nova arquitetura de rede com neurônios pulsados para a implementação de um sistema de reconhecimento automático do locutor. Para a realização dos testes foi utilizada a base de dados Speaker Recognition v1.0, do CSLU – Center for Spoken Language Understanding do Oregon Graduate Institute - E.U.A., contendo frases gravadas a partir de linhas telefônicas digitais. Para a etapa de classificação foi utilizada uma rede neural do tipo perceptron multicamada e os testes foram realizados no modo dependente e independente do texto. A viabilidade das Redes Neurais Pulsadas para o reconhecimento automático do locutor foi constatada, demonstrando que este paradigma neural é promissor para tratar as informações temporais do sinal de voz. / Pulsed Neural Networks have received a lot of attention from researchers. This work aims to verify the capability of this neural paradigm when applied to a speaker recognition task. After a description of the automatic speaker recognition and artificial neural networks fundamentals, a spike response model of neurons is tested. A novel neural network architecture based on this neuron model is proposed and used in a speaker recognition system. Text dependent and independent tests were performed using the Speaker Recognition v1.0 database from CSLU – Center for Spoken Language Understanding of Oregon Graduate Institute - U.S.A. A multilayer perceptron is used as a classifier. The Pulsed Neural Networks demonstrated its capability to deal with temporal information and the use of this neural paradigm in a speaker recognition task is promising.
132

Análise das concentrações energéticas no limiar entre fonemas vozeados e não-vozeados e suas implicações para fins de reconhecimento de locutores dependente do discurso / Analysis of energy cocentrations in the threshold between voiced and unvoiced phonemes and their implications for text-dependent speaker recognition

Ishizawa, William Habaro 19 February 2015 (has links)
Atualmente, diversos trabalhos e aplicações são desenvolvidos com foco na área de reconhecimento computacional de locutores. À medida que o interesse por diversas aplicações reais dentro dessa área emerge, principalmente em biometria, na qual a segurança e a eficácia são de extrema importância, torna-se cada vez mais necessário que estudos sejam feitos, na mesma proporção, visando avaliá-las. Desse modo, a proposta do presente trabalho é a de mensurar a acurácia de um sistema de reconhecimento de locutores baseado em características elementares, isto é, energias de sub-bandas de frequências, em associação com um classificador probabilístico, estudando a viabilidade de extraí-las das transições entre trechos vozeados e não-vozeados (TTVNV) dos sinais. Testes são realizados com diferentes quantidades de locutores e discurso fixado. A acurácia obtida nos testes variam de 20.18% a 92.53%. Os resultados obtidos são comparados e relatados, complementando as afirmações existentes na literatura sobre o uso das TTVNV com dados quantitativos. / Nowadays, many works and applications are developed focusing on computational speaker recognition. As the interest for several real applications within this area emerges, especially in biometrics, where the safety and the efficacy of the applications are extremely important, studies need to be developed in the same proportion, to evaluate the effectiveness of such approaches. Based on that, this work intends to measure the accuracy of a speaker recognition system that uses elementar features, i.e., sub-band frequency energies, associated with a probabilistic classifier, studying the viability of extracting them from the transition between voiced and unvoiced speech tags (TTVNV). Tests are carried out with different numbers of speakers and a text-dependent approach. The accuracy of the tests varies from 20.18% to 92.53%. The results are compared and reported, complementing the existent information on the use of TTVNV with quantitative data.
133

Verbos defectivos no português brasileiro: eles existem mesmo? / Defective verbs in Brazilian Portuguese: Do they really exist?

Oliveira, Klauber Renan Dutra de 19 September 2017 (has links)
Nesta dissertação, os verbos considerados defectivos pela normatividade são o objeto de estudo em questão. O objetivo consiste em averiguar se verbos, como banir, explodir, demolir, precaver, reaver, dentre outros são usados defectivamente pelos falantes. Para isso, nós analisamos e estudamos o conceito de defectividade em autores da tradição brasileira, como Cunha & Cintra (2008), Bechara (2009) e em autores que estudaram a defectividade no português brasileiro por meio de teorias linguísticas, como Nevins, Damulakis e Freitas (2014) e Maiden & ONeil (2010). Vimos que esses autores apontam causas da defectividade por questões fonológicas, morfológicas, semânticas e pragmáticas. Outros trabalhos que não tinham paradigmas verbais no português brasileiro foram fundamentais para o andamento desta pesquisa, como Baerman (2010) e Sims (2006). Dentre esses estudos analisados, procuramos abordar quais as causas identificadas pelos autores citados, dentre elas, encontramos a homofonia como um senso comum. Tratamos nossos dados dentro da teoria da gramática gerativa, como a teoria da otimalidade. Durante a nossa pesquisa, vimos que o fenômeno da defectividade ocorre de maneira desigual, pois, em alguns paradigmas verbais, falta apenas a primeira pessoa do singular do presente, enquanto em outras o paradigma apresenta apenas as formas arrizotônicas. Reanalisamos dados de pesquisas anteriores do ponto de vista fonológico, morfológico e semântico para mostrar alguns problemas nelas. Nosso estudo foi motivado porque encontramos dados os quais mostram que verbos ditos defectivos pela gramática tradicional são conjugados plenamente pelos falantes nativos. Logo, a fim de averiguar se o falante usa os mesmos verbos defectivamente como a gramática, fizemos alguns experimentos para isso. Os nossos resultados mostraram indícios de que a defectividade é vista de uma forma diferente pelo falante, pois há uma incompatibilidade parcial entre o que a gramática diz e entre o que o falante usa. / The verbs considered defective by normativity are the object of study in this thesis. The goal is to verify if verbs such as: ban, explode, demolish, preclude, retrieve, amongst others are used by the speakers defectively. To do so, we analyzed and studied the defectiveness concept in authors from the Brazilian tradition, such as Cunha & Cintra (2008), Bechara (2009), and in authors that have studied defectiveness in Brazilian Portuguese through other linguistic theories, such as Nevins, Damulakis and Freitas (2014) and Maiden & ONeil (2010). We have perceived that these authors point out reasons for defectiveness for phonologic, morphologic, semantic and pragmatic reasons. Other works that did not have Brazilian Portuguese verbal paradigms were fundamental for the development of this research, such as Baerman (2010) and Sims (2006). Amid the analyzed studies, we sought to approach the identified causes mentioned by the quoted authors, among them, we found the homophony as a common ground. We approached our data within the generative grammar theory, as well as the optimality theory. During our research, we saw that the defectiveness phenomenon happens in an uneven way because, in some verbal paradigms, there is only the present first-person singular missing, while in others, the paradigm presents only the forms with no stressed root. We reanalyzed the data from previous researches from the phonologic, morphologic and semantic point of view to present some of their problems. Our study was motivated by finding data that showed us that some verbs considered defective by the traditional grammar are fully conjugated by native speakers. Therefore, with the intent to verify if the speaker uses the same verbs defectively as the grammar, we made some experiments. Our results showed indications that the defectiveness is seen in a different way by the speaker, because there is a partial incompatibility between what is said by the grammar and what is used by the speaker.
134

Give me FAVE : Fault Analysis for Vibration in Electronics

Aljaderi, Maythem, Tang, Jocke, Mohammadi, Mohammad January 2012 (has links)
Ericsson har haft ett problem som påverkar deras mikrovågsradio. Detta problem handlar i grunden om mekaniska störningar som påverkar dataöverföringen mellan två radioenheter. Dessa störningar resulterar i bitfel på grund av olika orsaker. Dessa orsaker undersöks i projektet, för att i senare skede kunna förbättra precisionen av dataöverföringen. Genom att skicka signalerna med olika frekvenser på ett automatiserat och mer noggrant sätt, ökar möjligheten att testa radion i fler miljöer samtidigt möjligheten av att täcka ett så stort frekvensområde som möjligt ges. Arbetet är en blandning av elektronik, mekanik, akustik och programmering. Tanken är att den nya mätmetoden som presenteras skall vara automatisk och mjukvarustyrd. Även manuell styrning skall vara möjlig. Arbetet har bestått av forskning, marknadsskanning och kontakt med personer som är involverade inom området, detta för att hitta det bästa sättet att utveckla en ny felsökningsmetod.Med hjälp av olika testkörningar som studeras noggrant kommer förståelsen för ovan nämnda störningar att öka, vilket förhoppningsvis hjälper oss att hitta olika sätt att hantera dessa störningar i enskilda komponenter samt konstruktionen i sin helhet. Det finns flera förslag på alternativ, men genom ökad förståelse och kunskap inom området har det visat sig att det lämpligaste alternativet är att använda en shaker och en speaker som sändare och ett piezoelement som givare. Detta piezoelement tillsammans med en förstärkare mäter signalerna och övervakas med ett oscilloskop. Shakern och speakern drivs av en signalgenerator via en förstärkare. Alla dessa instrument styrs via ett styrprogram som är programmerad i LabVIEW. Styrprogrammets uppgift är att skanna över ett bestämt frekvensintervall med en konstant amplitud. Givaren mäter dessa signaler och sparar till en textfil. Denna information är viktig för att finna resonansfrekvenser och även övervaka den verkliga utsignalen som kommer fram till testobjektet.Detta arbete kommer förhoppningsvis att vara av stor betydelse för utvecklingen av nya produkter och kan bli ett användbart verktyg för andra ingenjörer inom Ericsson i framtiden. / Program: Utvecklingsingenjör
135

Sistemas de adaptação ao locutor utilizando autovozes. / Speaker adaptation system using eigenvoices.

Borges, Liselene de Abreu 20 December 2001 (has links)
O presente trabalho descreve duas técnicas de adaptação ao locutor para sistemas de reconhecimento de voz utilizando um volume de dados de adaptação reduzido. Regressão Linear de Máxima Verossimilhança (MLLR) e Autovozes são as técnicas trabalhadas. Ambas atualizam as médias das Gaussianas dos modelos ocultos de Markov (HMM). A técnica MLLR estima um grupo de transformações lineares para os parâmetros das medias das Gaussianas do sistema. A técnica de Autovozes baseia-se no conhecimento prévio das variações entre locutores. Para obtermos o conhecimento prévio, que está contido nas autovozes, utiliza-se a análise em componentes principais (PCA). Fizemos os testes de adaptação das médias em um sistema de reconhecimento de voz de palavras isoladas e de vocabulário restrito. Contando com um volume grande de dados de adaptação (mais de 70% das palavras do vocabulário) a técnica de autovozes não apresentou resultados expressivos com relação aos que a técnica MLLR apresentou. Agora, quando o volume de dados reduzido (menos de 15% das palavras do vocabulário) a técnica de Autovozes apresentou-se superior à MLLR. / This present work describe two speaker adaptation technique, using a small amount of adaptation data, for a speech recognition system. These techniques are Maximum Likelihood Linear Regression (MLLR) and Eigenvoices. Both re-estimates the mean of a continuous density Hidden Markov Model system. MLLR technique estimates a set of linear transformations for mean parameters of a Gaussian system. The eigenvoice technique is based on a previous knowledge about speaker variation. For obtaining this previous knowledge, that are retained in eigenvoices, it necessary to apply principal component analysis (PCA). We make adaptation tests over an isolated word recognition system, restrict vocabulary. If a large amount of adaptation data is available (up to 70% of all vocabulary) Eigenvoices technique does not appear to be a good implementation if compared with the MLLR technique. Now, when just a small amount of adaptation data is available (less than 15 % of all vocabulary), Eigenvoices technique get better results than MLLR technique.
136

Crossing barriers : the influence of linguistic and cultural background on [I + verb] belief constructions in expressions of opinion

Zhao, Lucy January 2017 (has links)
How does cultural and linguistic background influence communication style? This topic is examined through the [I + verb] belief construct before the expression of an opinion. Since opinions carry inherent notions of speaker belief, these constructions may at first appear superfluous. However, [I + verb] forms may actually fulfill various pragmatic functions depending on prosodic variation. Unfortunately, there is little congruent data on universality vs. cross-linguistic variability of pragmatic-prosodic mappings (prosodic variation as a cue to pragmatic interpretation) of [I + verb] belief constructs before an opinion. Thus, a Proof of Concept perception test was first implemented, followed by a production task investigating the effect of sociolinguistic background on a speaker's frequency of usage for various [I + verb] forms in expressing opinions, and how this relates to perceived speaker confidence. Usage of various forms and functions of this construct was analyzed and compared between native Mandarin (CHI) and English (US) speakers, as well as EFL Mandarin speakers. The Proof of Concept task supported hypotheses overall, suggesting a possible universal pragmatic-prosodic mapping for [I + verb]. In addition, while as predicted sociolinguistic background did not have a significant effect on universality of pragmatic-prosodic mapping in terms of confidence rating, it did have an observable effect on semantic interpretation of 'speaker confidence', thus indicating that sociolinguistic background may play a role in influencing these interpretations. Results from the production task supported predictions that frequency of functional [I + verb] usage corresponded to culturally specific attitudes of each culture. Based on confidence rating calculations for each [I + verb] variation from pragmatic-prosodic mapping of the perception task, it was determined that Native US individuals were most confident in expressing self-opinions but least confident in expressing opinions of others whilst Native CHI individuals were most confident in expressing opinions of others and least confident in expressing self-opinion, with the EFL group in the US more closely mirroring the Native US group and the EFL group in China more closely mirroring the Native CHI group. Additionally, going against theories of previous research, Time immersed in a new L2 environment and L2 proficiency did not significantly influence performance. Through investigating pragmatic-prosodic mappings of [I + verb] forms vs. functions, this study aimed to demonstrate the bi-directional link between language, thought and culture. By understanding and familiarizing oneself with the root of pragmatic differences, there is hope to better understand the cause of cross-cultural miscommunications between native and foreign speakers in conversation and to minimize any such discrepancies in pragmatic knowledge and sociocultural norms.
137

Sistemas de adaptação ao locutor utilizando autovozes. / Speaker adaptation system using eigenvoices.

Liselene de Abreu Borges 20 December 2001 (has links)
O presente trabalho descreve duas técnicas de adaptação ao locutor para sistemas de reconhecimento de voz utilizando um volume de dados de adaptação reduzido. Regressão Linear de Máxima Verossimilhança (MLLR) e Autovozes são as técnicas trabalhadas. Ambas atualizam as médias das Gaussianas dos modelos ocultos de Markov (HMM). A técnica MLLR estima um grupo de transformações lineares para os parâmetros das medias das Gaussianas do sistema. A técnica de Autovozes baseia-se no conhecimento prévio das variações entre locutores. Para obtermos o conhecimento prévio, que está contido nas autovozes, utiliza-se a análise em componentes principais (PCA). Fizemos os testes de adaptação das médias em um sistema de reconhecimento de voz de palavras isoladas e de vocabulário restrito. Contando com um volume grande de dados de adaptação (mais de 70% das palavras do vocabulário) a técnica de autovozes não apresentou resultados expressivos com relação aos que a técnica MLLR apresentou. Agora, quando o volume de dados reduzido (menos de 15% das palavras do vocabulário) a técnica de Autovozes apresentou-se superior à MLLR. / This present work describe two speaker adaptation technique, using a small amount of adaptation data, for a speech recognition system. These techniques are Maximum Likelihood Linear Regression (MLLR) and Eigenvoices. Both re-estimates the mean of a continuous density Hidden Markov Model system. MLLR technique estimates a set of linear transformations for mean parameters of a Gaussian system. The eigenvoice technique is based on a previous knowledge about speaker variation. For obtaining this previous knowledge, that are retained in eigenvoices, it necessary to apply principal component analysis (PCA). We make adaptation tests over an isolated word recognition system, restrict vocabulary. If a large amount of adaptation data is available (up to 70% of all vocabulary) Eigenvoices technique does not appear to be a good implementation if compared with the MLLR technique. Now, when just a small amount of adaptation data is available (less than 15 % of all vocabulary), Eigenvoices technique get better results than MLLR technique.
138

Verbos defectivos no português brasileiro: eles existem mesmo? / Defective verbs in Brazilian Portuguese: Do they really exist?

Klauber Renan Dutra de Oliveira 19 September 2017 (has links)
Nesta dissertação, os verbos considerados defectivos pela normatividade são o objeto de estudo em questão. O objetivo consiste em averiguar se verbos, como banir, explodir, demolir, precaver, reaver, dentre outros são usados defectivamente pelos falantes. Para isso, nós analisamos e estudamos o conceito de defectividade em autores da tradição brasileira, como Cunha & Cintra (2008), Bechara (2009) e em autores que estudaram a defectividade no português brasileiro por meio de teorias linguísticas, como Nevins, Damulakis e Freitas (2014) e Maiden & ONeil (2010). Vimos que esses autores apontam causas da defectividade por questões fonológicas, morfológicas, semânticas e pragmáticas. Outros trabalhos que não tinham paradigmas verbais no português brasileiro foram fundamentais para o andamento desta pesquisa, como Baerman (2010) e Sims (2006). Dentre esses estudos analisados, procuramos abordar quais as causas identificadas pelos autores citados, dentre elas, encontramos a homofonia como um senso comum. Tratamos nossos dados dentro da teoria da gramática gerativa, como a teoria da otimalidade. Durante a nossa pesquisa, vimos que o fenômeno da defectividade ocorre de maneira desigual, pois, em alguns paradigmas verbais, falta apenas a primeira pessoa do singular do presente, enquanto em outras o paradigma apresenta apenas as formas arrizotônicas. Reanalisamos dados de pesquisas anteriores do ponto de vista fonológico, morfológico e semântico para mostrar alguns problemas nelas. Nosso estudo foi motivado porque encontramos dados os quais mostram que verbos ditos defectivos pela gramática tradicional são conjugados plenamente pelos falantes nativos. Logo, a fim de averiguar se o falante usa os mesmos verbos defectivamente como a gramática, fizemos alguns experimentos para isso. Os nossos resultados mostraram indícios de que a defectividade é vista de uma forma diferente pelo falante, pois há uma incompatibilidade parcial entre o que a gramática diz e entre o que o falante usa. / The verbs considered defective by normativity are the object of study in this thesis. The goal is to verify if verbs such as: ban, explode, demolish, preclude, retrieve, amongst others are used by the speakers defectively. To do so, we analyzed and studied the defectiveness concept in authors from the Brazilian tradition, such as Cunha & Cintra (2008), Bechara (2009), and in authors that have studied defectiveness in Brazilian Portuguese through other linguistic theories, such as Nevins, Damulakis and Freitas (2014) and Maiden & ONeil (2010). We have perceived that these authors point out reasons for defectiveness for phonologic, morphologic, semantic and pragmatic reasons. Other works that did not have Brazilian Portuguese verbal paradigms were fundamental for the development of this research, such as Baerman (2010) and Sims (2006). Amid the analyzed studies, we sought to approach the identified causes mentioned by the quoted authors, among them, we found the homophony as a common ground. We approached our data within the generative grammar theory, as well as the optimality theory. During our research, we saw that the defectiveness phenomenon happens in an uneven way because, in some verbal paradigms, there is only the present first-person singular missing, while in others, the paradigm presents only the forms with no stressed root. We reanalyzed the data from previous researches from the phonologic, morphologic and semantic point of view to present some of their problems. Our study was motivated by finding data that showed us that some verbs considered defective by the traditional grammar are fully conjugated by native speakers. Therefore, with the intent to verify if the speaker uses the same verbs defectively as the grammar, we made some experiments. Our results showed indications that the defectiveness is seen in a different way by the speaker, because there is a partial incompatibility between what is said by the grammar and what is used by the speaker.
139

Reconhecimento automático do locutor com redes neurais pulsadas. / Automatic speaker recognition using pulse coupled neural networks.

Antonio Pedro Timoszczuk 22 March 2004 (has links)
As Redes Neurais Pulsadas são objeto de intensa pesquisa na atualidade. Neste trabalho é avaliado o potencial de aplicação deste paradigma neural, na tarefa de reconhecimento automático do locutor. Após uma revisão dos tópicos considerados importantes para o entendimento do reconhecimento automático do locutor e das redes neurais artificiais, é realizada a implementação e testes do modelo de neurônio com resposta por impulsos. A partir deste modelo é proposta uma nova arquitetura de rede com neurônios pulsados para a implementação de um sistema de reconhecimento automático do locutor. Para a realização dos testes foi utilizada a base de dados Speaker Recognition v1.0, do CSLU – Center for Spoken Language Understanding do Oregon Graduate Institute - E.U.A., contendo frases gravadas a partir de linhas telefônicas digitais. Para a etapa de classificação foi utilizada uma rede neural do tipo perceptron multicamada e os testes foram realizados no modo dependente e independente do texto. A viabilidade das Redes Neurais Pulsadas para o reconhecimento automático do locutor foi constatada, demonstrando que este paradigma neural é promissor para tratar as informações temporais do sinal de voz. / Pulsed Neural Networks have received a lot of attention from researchers. This work aims to verify the capability of this neural paradigm when applied to a speaker recognition task. After a description of the automatic speaker recognition and artificial neural networks fundamentals, a spike response model of neurons is tested. A novel neural network architecture based on this neuron model is proposed and used in a speaker recognition system. Text dependent and independent tests were performed using the Speaker Recognition v1.0 database from CSLU – Center for Spoken Language Understanding of Oregon Graduate Institute - U.S.A. A multilayer perceptron is used as a classifier. The Pulsed Neural Networks demonstrated its capability to deal with temporal information and the use of this neural paradigm in a speaker recognition task is promising.
140

Forensic speaker analysis and identification by computer : a Bayesian approach anchored in the cepstral domain

Khodai-Joopari, Mehrdad, Information Technology & Electrical Engineering, Australian Defence Force Academy, UNSW January 2007 (has links)
This thesis advances understanding of the forensic value of the automatic speech parameters by addressing the following question: what is the potentiality of the speech cepstrum as a forensic-acoustic parameter? Despite many advances in automatic speech and speaker recognition, robust and unconstrained progress in technical forensic speaker identification has been partly impeded by our incomplete understanding of the interaction and relation between forensic phonetics and the techniques employed in state-of-the-art automatic speech and speaker recognition. The posed question underlies the recurrent and longstanding issue of acoustic parameterisation in the area of forensic phonetics, where 1) speaker identification often must be carried out under less than optimal conditions, and 2) views differ on the usefulness and trustworthiness of the formant frequency measurements. To this end, a new formulation for the forensic evaluation of speech data was derived which is effectively a spectral likelihood ratio with enhanced sensitivity to the local peaks of the formant structure of the speech spectrum of vowel sounds, while retaining the characteristics of the Bayesian framework. This new hybrid formula was used together with a novel approach, which is founded on a statistically-based matched-pairs technique to account for various levels of variation inherent in speech recordings, thereby providing a spectrally meaningful measure of variations between two speech spectra and hence the true worth of speech samples as forensic evidence. The experimental results are obtained based on a forensically-realistic database of a relatively large population of 297 native speakers of Japanese. In sum, the research conducted in this thesis is a major step forward in advancing the forensic-phonetic field which broadens the objective basis of the forensic speaker identification. Beyond advancing knowledge in the field, the semi data-independent nature of the new formula ultimately has great implications in technical forensic speaker identification. It also provides us with a valuable biometric tool with both academic and commercial potential in crime investigation in a field which is already suffering from the lack of adequate data.

Page generated in 0.0449 seconds