In speech communication systems such as voice-controlled systems, hands-free mobile telephones and hearing aids, the received signals are degraded by room reverberation and background noise. This degradation can reduce the perceived quality and intelligibility of the speech, and decrease the performance of speech enhancement and source localization. These problems are difficult to solve due to the colored and nonstationary nature of the speech signals, and features of the Room Impulse Response (RIR) such as its long duration and non-minimum phase. In this dissertation, we focus on two topics of speech enhancement and speaker localization in noisy reverberant environments. A two-stage speech enhancement method is presented to suppress both early and late reverberation in noisy speech using only one microphone. It is shown that this method works well even in highly reverberant rooms. Experiments under different acoustic conditions confirm that the proposed blind method is superior in terms of reducing early and late reverberation effects and noise compared to other well known single-microphone techniques in the literature. Time Difference Of Arrival (TDOA)-based methods usually provide the most accurate source localization in adverse conditions. The key issue for these methods is to accurately estimate the TDOA using the smallest number of microphones. Two robust Time Delay Estimation (TDE) methods are proposed which use the information from only two microphones. One method is based on adaptive inverse filtering which provides superior performance even in highly reverberant and moderately noisy conditions. It also has negligible failure estimation which makes it a reliable method in realistic environments. This method has high computational complexity due to the estimation in the first stage for the first microphone. As a result, it can not be applied in time-varying environments and real-time applications. Our second method improves this problem by introducing two effective preprocessing stages for the conventional Cross Correlation (CC)-based methods. The results obtained in different noisy reverberant conditions including a real and time-varying environment demonstrate that the proposed methods are superior compared to the conventional TDE methods. / Graduate / 0544 / 0984 / saeed.mosayyebpour@gmail.com

In speech communication systems such as voice-controlled systems, hands-free mobile telephones and hearing aids, the received signals are degraded by room reverberation and background noise. This degradation can reduce the perceived quality and intelligibility of the speech, and decrease the performance of speech enhancement and source localization. These problems are difficult to solve due to the colored and nonstationary nature of the speech signals, and features of the Room Impulse Response (RIR) such as its long duration and non-minimum phase. In this dissertation, we focus on two topics of speech enhancement and speaker localization in noisy reverberant environments. A two-stage speech enhancement method is presented to suppress both early and late reverberation in noisy speech using only one microphone. It is shown that this method works well even in highly reverberant rooms. Experiments under different acoustic conditions confirm that the proposed blind method is superior in terms of reducing early and late reverberation effects and noise compared to other well known single-microphone techniques in the literature. Time Difference Of Arrival (TDOA)-based methods usually provide the most accurate source localization in adverse conditions. The key issue for these methods is to accurately estimate the TDOA using the smallest number of microphones. Two robust Time Delay Estimation (TDE) methods are proposed which use the information from only two microphones. One method is based on adaptive inverse filtering which provides superior performance even in highly reverberant and moderately noisy conditions. It also has negligible failure estimation which makes it a reliable method in realistic environments. This method has high computational complexity due to the estimation in the first stage for the first microphone. As a result, it can not be applied in time-varying environments and real-time applications. Our second method improves this problem by introducing two effective preprocessing stages for the conventional Cross Correlation (CC)-based methods. The results obtained in different noisy reverberant conditions including a real and time-varying environment demonstrate that the proposed methods are superior compared to the conventional TDE methods. / Graduate / 2015-04-23 / 0544 / 0984 / saeed.mosayyebpour@gmail.com

Em vista da tentência de se criarem intefaces entre humanos e máquinas que cada vez mais permitam meios simples de interação, é natural que sejam realizadas pesquisas em técnicas que procuram simular o meio mais convencional de comunicação que os humanos usam: a fala. No sistema auditivo humano, a voz é automaticamente processada pelo cérebro de modo efetivo e fácil, também comumente auxiliada por informações visuais, como movimentação labial e localizacão dos locutores. Este processamento realizado pelo cérebro inclui dois componentes importantes que a comunicação baseada em fala requere: Detecção de Atividade de Voz (Voice Activity Detection - VAD) e Localização de Fontes Sonoras (Sound Source Localization - SSL). Consequentemente, VAD e SSL também servem como ferramentas mandatórias de pré-processamento em aplicações de Interfaces Humano-Computador (Human Computer Interface - HCI), como no caso de reconhecimento automático de voz e identificação de locutor. Entretanto, VAD e SSL ainda são problemas desafiadores quando se lidando com cenários acústicos realísticos, particularmente na presença de ruído, reverberação e locutores simultâneos. Neste trabalho, são propostas abordagens para tratar tais problemas, para os casos de uma e múltiplas fontes sonoras, através do uso de informações audiovisuais, explorando-se variadas maneiras de se fundir as modalidades de áudio e vídeo. Este trabalho também emprega um arranjo de microfones para o processamento de som, o qual permite que as informações espaciais dos sinais acústicos sejam exploradas através do algoritmo estado-da-arte SRP (Steered Response Power). Por consequência adicional, uma eficiente implementação em GPU do SRP foi desenvolvida, possibilitando processamento em tempo real do algoritmo. Os experimentos realizados mostram uma acurácia média de 95% ao se efetuar VAD de até três locutores simultâneos, e um erro médio de 10cm ao se localizar tais locutores. / Given the tendency of creating interfaces between human and machines that increasingly allow simple ways of interaction, it is only natural that research effort is put into techniques that seek to simulate the most conventional mean of communication humans use: the speech. In the human auditory system, voice is automatically processed by the brain in an effortless and effective way, also commonly aided by visual cues, such as mouth movement and location of the speakers. This processing done by the brain includes two important components that speech-based communication require: Voice Activity Detection (VAD) and Sound Source Localization (SSL). Consequently, VAD and SSL also serve as mandatory preprocessing tools for high-end Human Computer Interface (HCI) applications in a computing environment, as the case of automatic speech recognition and speaker identification. However, VAD and SSL are still challenging problems when dealing with realistic acoustic scenarios, particularly in the presence of noise, reverberation and multiple simultaneous speakers. In this work we propose some approaches for tackling these problems using audiovisual information, both for the single source and the competing sources scenario, exploiting distinct ways of fusing the audio and video modalities. Our work also employs a microphone array for the audio processing, which allows the spatial information of the acoustic signals to be explored through the stateof- the art method Steered Response Power (SRP). As an additional consequence, a very fast GPU version of the SRP is developed, so that real-time processing is achieved. Our experiments show an average accuracy of 95% when performing VAD of up to three simultaneous speakers and an average error of 10cm when locating such speakers.

Le secteur des transports ferroviaires en France est marqué par un dynamisme lié notamment à l'essor du réseau à grande vitesse et à la réimplantation du tramway dans de nombreuses agglomérations. Dans ce contexte, la réduction des nuisances sonores apparaît comme un enjeu majeur pour son développement. Afin d'agir efficacement à la source, il est indispensable d'identifier et d'étudier précisément les sources responsables de ces nuisances au passage des véhicules. Parmi les approches possibles, les antennes microphoniques et les traitements associés sont particulièrement adaptés à la caractérisation des sources ponctuelles mobiles, omnidirectionnelles et décorrélées.Pour les vitesses inférieures à 300 km/h, le bruit de roulement constitue la source principale du bruit ferroviaire ; il résulte du rayonnement acoustique des éléments tels que les roues, le rail et les traverses. Le rail, dont la contribution au bruit de roulement est prépondérante aux moyennes fréquences (entre 500 He et 1000 Hz environ), est une source étendue et cohérente pour laquelle les principes classiques de traitement d'antenne ne sont pas adaptés.La méthode de caractérisation proposée dans cette thèse est une méthode inverse d'optimisation paramétrique utilisant les signaux acoustiques issus d'une antenne microphonique. Les paramètres inconnus d'un modèle vibro-acoustique sont estimés par minimisation d'un critère des moindres carrés sur les matrices spectrales mesurée et modélisée au niveau de l'antenne. Dans le modèle vibro-acoustique, le rail est assimilé à un monopôle cylindrique dont la distribution longitudinale d'amplitude est liée à celle des vitesses vibratoires. Pour le calcul de ces vitesses, les différents modèles proposés mettent en évidence des ondes vibratoires se propageant dans le rail de part et d'autre de chaque excitation. Chacune de ces ondes est caractérisée par une amplitude au niveau de l'excitation, un nombre d'onde structural réel et une atténuation. Ces paramètres sont estimés par minimisation du critère, puis utilisés pour reconstruire le champ acoustique.Dans un premier temps, des simulations sont réalisées pour juger des performances de la méthode proposée, dans le cas d'excitations ponctuelles verticales. En particulier, sa robustesse est testée en présence de bruit ou d'incertitudes sur les paramètres supposés connus du modèle. Les effets de l'utilisation de modèles dégradés sont également étudiés. Concernant l'estimation des amplitudes, les résultats ont montré que la méthode est particulièrement robuste et efficace pour les excitations les plus proches de l'antenne. En revanche, pour l'estimation des autres paramètres, les performances sont supérieures pour les positions d'antenne excentrées. De manière générale, le nombre d'onde est correctement estimé sur l'ensemble des fréquences étudiées. Dans les cas à faible atténuation, un traitement classique par formation de voies en ondes planes suffit. En ce qui concerne l'estimation de l'atténuation, la faible sensibilité du critère limite l'efficacité de la méthode proposée.Enfin, certains résultats obtenus à partir des simulations ont été vérifiés lors de mesures in situ. L'excitation d'un rail expérimental par un marteau de chocs a tout d'abord permis de valider le modèle vibratoire pour la flexion verticale. Pour tester la méthode d'optimisation paramétrique, le rail a également été excité verticalement à l'aide d'un pot vibrant. Les principaux résultats des simulations ont été retrouvés, et des comportements particuliers relatifs à la présence de plusieurs ondes dans le rail ont été observés, ouvrant des perspectives de généralisation du modèle vibratoire utilisé. / In France, railway transport has been boosted by the expansion of the high-speed rail service and the resurgent implantation of tram networks in many city centers. In this context, the reduction of noise pollution becomes a crucial issue for its development. In order to directly act on the source area, it is necessary to precisely identify and study the sources responsible for this nuisance at train pass-by. Among all the potential approaches, microphone arrays and related signal processing techniques are particularly adapted to the characterization of omnidirectional and uncorrelated moving point sources. For speeds up to 300 km/h, rolling noise is the main railway noise source. It arises from the acoustic radiation of various elements such as wheels, rail or sleepers. The rail, which mainly contributes to rolling noise at mid-frequencies (from 500 Hz to 1000 Hz approximately), is an extended coherent source for which classical array processing methods are inappropriate. The characterization method proposed in this thesis is an inverse parametric optimization method that uses the acoustical signals measured by a microphone array. The unknown parameters of a vibro-acoustical model are estimated through the minimization of a least square criterion applied to the entries of the measured and modelled spectral matrices. In this vibro-acoustical model, the rail is considered as a cylindrical monopole whose lengthwise amplitude distribution is obtained from the vibratory velocity one. The different models proposed to obtain this velocity highlight the propagation of vibration waves towards both sides of every forcing point. Each wave is characterized by an amplitude at the forcing point, a real structural wavenumber and a decay rate. These parameters are estimated by the minimization of the least square criterion, and are then used in the vibro-acoustical model to rebuild the acoustical field radiated by the rail. First, simulations are performed in order to appraise the performances of the proposed method, in the case of vertical point excitations. In particular, its robustness to additive noise and to uncertainties in the model parameters that are supposed to be known is tested. The effect of using simplified models is also investigated. Results show that the method is efficient and robust for the amplitude estimation of the nearest contacts to the array. On the other hand, the estimation of the other parameters is improved when the array is shifted away from the contact points. The wavenumber is generally well estimated over the entire frequency range, and when the decay rate is low, a single beamforming technique may be sufficient. Concerning the decay rate estimation, the efficiency of the method is limited by the low sensitivity of the criterion. At last, measurements are performed in order to verify some results obtained from the simulations. The vibratory model is first validated for the vertical flexural waves trough the use of an impact hammer. Then, the parametric optimization method is tested by the vertical excitation of the rail with a modal shaker. The main simulation results are found, and some particular behavior due to other waves existing in the rail can be observed, opening the perspective of a generalized method including more complex vibratory modelings.

Em vista da tentência de se criarem intefaces entre humanos e máquinas que cada vez mais permitam meios simples de interação, é natural que sejam realizadas pesquisas em técnicas que procuram simular o meio mais convencional de comunicação que os humanos usam: a fala. No sistema auditivo humano, a voz é automaticamente processada pelo cérebro de modo efetivo e fácil, também comumente auxiliada por informações visuais, como movimentação labial e localizacão dos locutores. Este processamento realizado pelo cérebro inclui dois componentes importantes que a comunicação baseada em fala requere: Detecção de Atividade de Voz (Voice Activity Detection - VAD) e Localização de Fontes Sonoras (Sound Source Localization - SSL). Consequentemente, VAD e SSL também servem como ferramentas mandatórias de pré-processamento em aplicações de Interfaces Humano-Computador (Human Computer Interface - HCI), como no caso de reconhecimento automático de voz e identificação de locutor. Entretanto, VAD e SSL ainda são problemas desafiadores quando se lidando com cenários acústicos realísticos, particularmente na presença de ruído, reverberação e locutores simultâneos. Neste trabalho, são propostas abordagens para tratar tais problemas, para os casos de uma e múltiplas fontes sonoras, através do uso de informações audiovisuais, explorando-se variadas maneiras de se fundir as modalidades de áudio e vídeo. Este trabalho também emprega um arranjo de microfones para o processamento de som, o qual permite que as informações espaciais dos sinais acústicos sejam exploradas através do algoritmo estado-da-arte SRP (Steered Response Power). Por consequência adicional, uma eficiente implementação em GPU do SRP foi desenvolvida, possibilitando processamento em tempo real do algoritmo. Os experimentos realizados mostram uma acurácia média de 95% ao se efetuar VAD de até três locutores simultâneos, e um erro médio de 10cm ao se localizar tais locutores. / Given the tendency of creating interfaces between human and machines that increasingly allow simple ways of interaction, it is only natural that research effort is put into techniques that seek to simulate the most conventional mean of communication humans use: the speech. In the human auditory system, voice is automatically processed by the brain in an effortless and effective way, also commonly aided by visual cues, such as mouth movement and location of the speakers. This processing done by the brain includes two important components that speech-based communication require: Voice Activity Detection (VAD) and Sound Source Localization (SSL). Consequently, VAD and SSL also serve as mandatory preprocessing tools for high-end Human Computer Interface (HCI) applications in a computing environment, as the case of automatic speech recognition and speaker identification. However, VAD and SSL are still challenging problems when dealing with realistic acoustic scenarios, particularly in the presence of noise, reverberation and multiple simultaneous speakers. In this work we propose some approaches for tackling these problems using audiovisual information, both for the single source and the competing sources scenario, exploiting distinct ways of fusing the audio and video modalities. Our work also employs a microphone array for the audio processing, which allows the spatial information of the acoustic signals to be explored through the stateof- the art method Steered Response Power (SRP). As an additional consequence, a very fast GPU version of the SRP is developed, so that real-time processing is achieved. Our experiments show an average accuracy of 95% when performing VAD of up to three simultaneous speakers and an average error of 10cm when locating such speakers.

Em vista da tentência de se criarem intefaces entre humanos e máquinas que cada vez mais permitam meios simples de interação, é natural que sejam realizadas pesquisas em técnicas que procuram simular o meio mais convencional de comunicação que os humanos usam: a fala. No sistema auditivo humano, a voz é automaticamente processada pelo cérebro de modo efetivo e fácil, também comumente auxiliada por informações visuais, como movimentação labial e localizacão dos locutores. Este processamento realizado pelo cérebro inclui dois componentes importantes que a comunicação baseada em fala requere: Detecção de Atividade de Voz (Voice Activity Detection - VAD) e Localização de Fontes Sonoras (Sound Source Localization - SSL). Consequentemente, VAD e SSL também servem como ferramentas mandatórias de pré-processamento em aplicações de Interfaces Humano-Computador (Human Computer Interface - HCI), como no caso de reconhecimento automático de voz e identificação de locutor. Entretanto, VAD e SSL ainda são problemas desafiadores quando se lidando com cenários acústicos realísticos, particularmente na presença de ruído, reverberação e locutores simultâneos. Neste trabalho, são propostas abordagens para tratar tais problemas, para os casos de uma e múltiplas fontes sonoras, através do uso de informações audiovisuais, explorando-se variadas maneiras de se fundir as modalidades de áudio e vídeo. Este trabalho também emprega um arranjo de microfones para o processamento de som, o qual permite que as informações espaciais dos sinais acústicos sejam exploradas através do algoritmo estado-da-arte SRP (Steered Response Power). Por consequência adicional, uma eficiente implementação em GPU do SRP foi desenvolvida, possibilitando processamento em tempo real do algoritmo. Os experimentos realizados mostram uma acurácia média de 95% ao se efetuar VAD de até três locutores simultâneos, e um erro médio de 10cm ao se localizar tais locutores. / Given the tendency of creating interfaces between human and machines that increasingly allow simple ways of interaction, it is only natural that research effort is put into techniques that seek to simulate the most conventional mean of communication humans use: the speech. In the human auditory system, voice is automatically processed by the brain in an effortless and effective way, also commonly aided by visual cues, such as mouth movement and location of the speakers. This processing done by the brain includes two important components that speech-based communication require: Voice Activity Detection (VAD) and Sound Source Localization (SSL). Consequently, VAD and SSL also serve as mandatory preprocessing tools for high-end Human Computer Interface (HCI) applications in a computing environment, as the case of automatic speech recognition and speaker identification. However, VAD and SSL are still challenging problems when dealing with realistic acoustic scenarios, particularly in the presence of noise, reverberation and multiple simultaneous speakers. In this work we propose some approaches for tackling these problems using audiovisual information, both for the single source and the competing sources scenario, exploiting distinct ways of fusing the audio and video modalities. Our work also employs a microphone array for the audio processing, which allows the spatial information of the acoustic signals to be explored through the stateof- the art method Steered Response Power (SRP). As an additional consequence, a very fast GPU version of the SRP is developed, so that real-time processing is achieved. Our experiments show an average accuracy of 95% when performing VAD of up to three simultaneous speakers and an average error of 10cm when locating such speakers.

The theme of the master´s thesis is design a microphone preamplifier which uses vacuum tube as a main amplifying element. Part of the work is theoretical assumptions for the optimal design and implementation of peripheral involvement, comparing the properties of components used and the resulting parameters of modeled device. The thesis also includes a proposal of the source unit for supplying all parts of the preamplifier. The results are accompanied by circuit simulations and laboratory measurements of the main parameters of the designed device.

<p>Главни задатак истраживања приказаног у дисертацији је развој модела,<br />заснованог на алгоритмима машинског учења, који описује сложени<br />утицај мерног система на користан, експериментални сигнал са циљем<br />његове елиминације. Студија случаја је широко распрострањена<br />фотоакустична, трансмисиона мерна метода са ћелијом минималне<br />запремине. Мултидисциплинарност и комплексност проблема одредили<br />су следеће кораке у методологији решења: 1) развој софтвера за<br />генерисање симулираних експерименталних података, 2) развој<br />регресионог модела заснованог на трослојној неуронској мрежи, за<br />прецизну и поуздану карактеризацију детектора која се извршава у<br />реалном времену, 3) развој класификационог модела заснованог на<br />неуронској мрежи једноставне структуре за прецизну и поуздану<br />предикцију типа коришћеног детектора која се извршава у реалном<br />времену, 4) спрезање регресионог и класификационог модела уз развој<br />додатног софтвера за прилагођење модела стварном експерименту. На<br />овај начин заокружен је програмски оквир који извршава сложени задатак<br />издвајања &ldquo;правог&rdquo; сигнала oд изобличеног експерименталног сигнала<br />без ангажовања истраживача, односно извршава аутокорекцију.<br />Тестирање је извршено на више различитих детектора и више<br />различитих материјала у фотоаксустичном експерименту. Применом<br />развијеног програмског оквира конкурентност експерименталне технике<br />је знатно порасла: повећана је тачност и поузданост, проширен је мерни<br />опсег и смањено време обраде резултата мерења.</p> / <p>Glavni zadatak istraživanja prikazanog u disertaciji je razvoj modela,<br />zasnovanog na algoritmima mašinskog učenja, koji opisuje složeni<br />uticaj mernog sistema na koristan, eksperimentalni signal sa ciljem<br />njegove eliminacije. Studija slučaja je široko rasprostranjena<br />fotoakustična, transmisiona merna metoda sa ćelijom minimalne<br />zapremine. Multidisciplinarnost i kompleksnost problema odredili<br />su sledeće korake u metodologiji rešenja: 1) razvoj softvera za<br />generisanje simuliranih eksperimentalnih podataka, 2) razvoj<br />regresionog modela zasnovanog na troslojnoj neuronskoj mreži, za<br />preciznu i pouzdanu karakterizaciju detektora koja se izvršava u<br />realnom vremenu, 3) razvoj klasifikacionog modela zasnovanog na<br />neuronskoj mreži jednostavne strukture za preciznu i pouzdanu<br />predikciju tipa korišćenog detektora koja se izvršava u realnom<br />vremenu, 4) sprezanje regresionog i klasifikacionog modela uz razvoj<br />dodatnog softvera za prilagođenje modela stvarnom eksperimentu. Na<br />ovaj način zaokružen je programski okvir koji izvršava složeni zadatak<br />izdvajanja &ldquo;pravog&rdquo; signala od izobličenog eksperimentalnog signala<br />bez angažovanja istraživača, odnosno izvršava autokorekciju.<br />Testiranje je izvršeno na više različitih detektora i više<br />različitih materijala u fotoaksustičnom eksperimentu. Primenom<br />razvijenog programskog okvira konkurentnost eksperimentalne tehnike<br />je znatno porasla: povećana je tačnost i pouzdanost, proširen je merni<br />opseg i smanjeno vreme obrade rezultata merenja.</p> / <p>The main task of the research presented in this dissertation is the development<br />of the model based on machine learning algorithms, which describes the<br />complex influence of the measuring system on a useful, experimental signal,<br />with the aim of the elimination of this influence. The case study is a widespread<br />photoacoustic, transmission measurement method with minimum volume cell<br />configuration. Multidisciplinarity and complexity of the problem determined the<br />following steps in the solution methodology: 1) development of the software for<br />generating simulated experimental data, 2) development of the regression<br />model based on a three-layer neural network, for precise and reliable<br />characterization of detectors, performed in real time, 3) development of the<br />classification model based on a neural network of simple structure for precise<br />and reliable prediction of the type of detector in use, performed in real time, 4)<br />coupling of the regression and the classification model with the development<br />of additional software for adjustment of the model to a real experiment. In this<br />way, the program framework is completed, which performs the complex task<br />of extracting the &quot;true&quot; signal from the distorted experimental signal without the<br />involvement of researchers, performing, thus, the autocorrection. Testing was<br />performed on several different detectors and several different materials in a<br />photoacoustic experiment. With the application of the developed software<br />framework, the competitiveness of the experimental technique has<br />significantly increased: the accuracy and the reliability have been increased,<br />the measurement range has been expanded and the processing time of<br />measurement results has been reduced.</p>

The aim of the master thesis is to sum up theoretical information about beamforming methods of microphone arrays and to verify their functionality. At the beginning of this work there are simulated different varietes of linear uniform and nonuniform microphone arrays and circular arrays. The results are verificated by a practical measurement in ideal conditions. Then I will focuse on implementation of the DAS(Delay And Sum), SAB(Sub Array Beamforming), CDB(Constant Directivity Beamforming), CDB-CA(CDB-Circular Arrays) beamformer including theoretical and practical verification of the functionality in ideal conditions. At the end of this thesis are all beamforming methods compared with each other at SNR(signal to Noise Ratio) and directivity parameters.

The aim of this work is to implement digital filters into programmable gate array. The work also includes a description of the MEMS technology, including comparisons with the technology of MEMS microphones from various manufacturers. Another part is devoted to the Sigma-delta modulation. The main section is the design and implementation of digital CIC and FIR filters for signal processing of digital microphone, including simulation and verification of properties of the proposed filter in Matlab.