|
|
Computational modelling of voice propagation around the human head using finite element method
Ryšavý, Antonín ; Hájek, Petr (referee) ; Švancara, Pavel (advisor)
In the first part of this master's thesis there is briefly presented the biomechanics of the human voice creation and an overview of the hitherto published computational models of the vocal tract and dissemination of the acoustic waves around the human head. The second part of the thesis deals with the creation of the computational model of a dissemination of the acoustic waves through vocal tract and further into a space around the head during the pronouncing of the Czech vowel /a:/. The vocal tract is excited by a harmonic signal in the place of vocal chords. On the surface of the vocal tract and the part of the head including hair and skin there is defined an acoustic absorption. The dissemination of the acoustic waves in the vocal tract, in the near field around the mouth, in the area around the head and in the points on the cheeks is detailed mapped. The dissemination of the acoustic waves is analyzed in the points where the speech microphones are placed. Acoustic pressure dependence on frequency, transmission functions between defined points and the acoustic pressure amplitudes depending on the distance from the mouth are obtained. In particular, the frequency distortion of the spectra is observed at the points indicated. Furthermore, the radiation impedance in the mouth area is evaluated. The results obtained are compared with the results of the hitherto published experimental measurements and can be used for the exact measurement of human voice or for the frequency correction of the microphones during the scann of the speech and sing placced in the analyzed points.
|
|
|
Using of Transfer Matrix Method for Modal Characteristics of Vocal Tract
Meisner, Patrik ; Švancara, Pavel (referee) ; Hájek, Petr (advisor)
The first part of the bachelor thesis describes anatomy of the vocal tract, voice creation and its modification. Some pathologies are listed at the end of the first part. The contents of second part are previously published computational models of the vocal tract followed by the method of transfer matrices for the Czech vowel [u:]. Modal analysis is performed by the transfer matrix method. The output are first three natural frequencies. Obtained frequencies are compared to values in the literature and to values calculated by the finite element method.
|
|
|
Creating a computational model of the human vocal tract
Freiwald, Michal ; Hájek, Petr (referee) ; Švancara, Pavel (advisor)
The research part of this bachelor’s thesis consists of a brief introduction to the human respiratory system and its subsidiary vocal subsystem along with a summary of basic phonation theories, voice disorders and published computational models of the human vocal tract. The experimental part engages in the making of the computational model itself, set to pronunciation of the vowel /a:/ in a woman vocal tract, on which, using finite element method, some of the basic acoustic analyses are performed, such as modal analysis or harmonic analysis. Calculated formants correspond with the values published in literature. Several different methods were analyzed while computing harmonic response. The most complex and the most time-consuming method, using infinite elements, also proved to be the most precise one. Thesis gives a decent comparison of the precision and complexity between the used methods.
|
|
|
Stress recognition from speech signal
Staněk, Miroslav ; Přibil, Jiří (referee) ; Tučková,, Jana (referee) ; Sigmund, Milan (advisor)
Předložená disertační práce se zabývá vývojem algoritmů pro detekci stresu z řečového signálu. Inovativnost této práce se vyznačuje dvěma typy analýzy řečového signálu, a to za použití samohláskových polygonů a analýzy hlasivkových pulsů. Obě tyto základní analýzy mohou sloužit k detekci stresu v řečovém signálu, což bylo dokázáno sérií provedených experimentů. Nejlepších výsledků bylo dosaženo pomocí tzv. Closing-To-Opening phase ratio příznaku v Top-To-Bottom kritériu v kombinaci s vhodným klasifikátorem. Detekce stresu založená na této analýze může být definována jako jazykově i fonémově nezávislá, což bylo rovněž dokázáno získanými výsledky, které dosahují v některých případech až 95% úspěšnosti. Všechny experimenty byly provedeny na vytvořené české databázi obsahující reálný stres, a některé experimenty byly také provedeny pro anglickou stresovou databázi SUSAS.
|
|
|
Analysis of Formant Areas During Sung Vowels
Kukosa, František ; Švancara, Pavel (referee) ; Jirásek, Ondřej (advisor)
The study analyzes the formant zones of male voices in singing vowels with main focus on opera, chest and head vocal registers. The results of this work are based on the recordings of five singers, taken in an anechoic chamber. For the analysis of these recordings in the frequency domain were used the FFT algorithm and the LPC analysis. The results describe the nature of the individual registers and show the importance of the singing formants for singers. Especially for opera vocal register where the vocal formants, along with a natural vibrato technique, are very important.
|
|
|
Diagnosing Parkinson's disease from analysis of speech recording
Vymlátil, Petr ; Trzos, Michal (referee) ; Lněnička, Jakub (advisor)
This thesis is focused on diagnosing Parkinson’s disease from analysis of speech recording. Introduction of this work deals with description of voice production mechanism, it’s basic qualities and influence of hypokinetic dysarthria on speech. In next chapter, there is described voice signal and some methods of it’s preprocessing. Next part continues dealing with description of chosen individual symptoms, which are needed for PD diagnosing, followed by definition of chosen reduction methods and classifiers. There is a comparison of classify succes of naive bayes classifier, depending on chosen reduction method in last chapter of this work.
|
|
|
Computational modelling of function of the human vocal tract
Ryšavý, Antonín ; Hájek, Petr (referee) ; Švancara, Pavel (advisor)
In the first part of this bachelor's thesis is a brief summary of the biomechanics of the creation of the human voice and an overview of the published computational models of the vocal tract and the area around the head. The second part deals with the computational models of the human vocal tract set to the pronouncing the Czech vowels /a:/ and /i:/ with using the method of transfer matrices and the finite element method. By these methods is perform modal and harmonic analysis. Are investigated the natural frequencies and own vibration shapes of both vowels and course of sound pressure in a specific areas of the vocal tract. The method of transfer matrices is highly depend on the geometry of the tract, particularly on the density of the reference sections and its results in this thesis do not completely agree with the results in the literature. Finite element method is more accurate and its results agree well with results reported in the literature, but the opposite of the transfer matrices method is significantly time consuming. Method of the transfer matrices is more suitable for a large number of calculations or tuning certain parameters. Models created in this bachelor's thesis can serve for the analyse of pathology of voice production, eventually for prediction of surgical procedures in the area of the vocal tract.
|
|
|
Finite element modelling of voice propagation through the vocal tract and around the human head
Tomeček, Vojtěch ; Pellant, Karel (referee) ; Švancara, Pavel (advisor)
This diploma thesis deals with creating of finite element model for the analysis of acustic waves radiation through human vocal tract and through the head surrounding space. The thesis contains a short overview of the anatomy of relevant parts and a short overview of formerly published models relevant to the topic. Then the process of creating the 3D model of human head, including the vocal tract, based on CT scans, and the mesh itself follows. This is succeeded by the computational solution taking in the acount the acoustic absorption of the vocal tract walls and radiation into the open space. The results gained are compared to relevant literature on this topic. The changes in frequency spectra in specific nodes of the mesh are reviewed and compared to relevant literature on the topic. The results could be used as basis of eventual frequention corrections of microphones used for voice diagnosis or registration in general, eg of speach or singing.
|
|
|
LPC Speech Coding
Zapletal, Ondřej ; Kyselý, František (referee) ; Rajmic, Pavel (advisor)
The contents of the thesis "LPC speech coding" are studies of this method of a parametric source coding, explanation of mathematical procedures that are used in it (linear prediction, autocorrelation, Levinson-Durbin algorithm, transfer to a form suitable for transmission, Chebyshev root searching polynomial method) and acquaintance with the signification and application of that method in real speech encoders. The task of the original project of this thesis is a description and simulation of a simple speech encoder based on LPC, which transforms a real speech signal into a bit flow, which contains all of the significant parameters for its backward reconstruction (LSF coefficients, pitch period, excitation level, voice detection - AMDF method). One part of this thesis is a discussion about currently used speech encoders.
|
|
|
Parkinson disease diagnosis using speech signal analysis
Karásek, Michal ; Smékal, Zdeněk (referee) ; Mekyska, Jiří (advisor)
The thesis deals with the recognition of Parkinson's disease from the speech signal. The first part refers to the principles of speech signals and speech signals by patients suffering from Parkinson's disease. Further, it continues to describe the issues of speech signals processing, basic symptoms used for diagnosis of Parkinson's disease (e. g. VAI, VSA, FCR, VOT etc.) and reduction of these symptoms. The next part focuses on a block diagram of the program for the diagnosis of Parkinson's disease. The main objective of this thesis is comparison of two methods of feature selection (mRMR and SFFS). For classification have selected two different methods were used. The first method is classification kNN and second method of classification is Gaussian mixture model (GMM).
|
guest ::
