A computational model of prosody for Yorùbá text-to-speech synthesis

This work examines prosody modelling for the Standard Yorùbá (SY) language in the context of computer text-to-speech synthesis applications. The thesis of this research is that it is possible to develop a practical prosody model by using appropriate computational tools and techniques which combines acoustic data with an encoding of the phonological and phonetic knowledge provided by experts. Our prosody model is conceptualised around a modular holistic framework. The framework is implemented using the Relational Tree (R-Tree) techniques (Ehrich and Foith, 1976). R-Tree is a sophisticated data structure that provides a multi-dimensional description of a waveform. A Skeletal Tree (S-Tree) is first generated using algorithms based on the tone phonological rules of SY. Subsequent steps update the S-Tree by computing the numerical values of the prosody dimensions. To implement the intonation dimension, fuzzy control rules where developed based on data from native speakers of Yorùbá. The Classification And Regression Tree (CART) and the Fuzzy Decision Tree (FDT) techniques were tested in modelling the duration dimension. The FDT was selected based on its better performance. An important feature of our R-Tree framework is its flexibility in that it facilitates the independent implementation of the different dimensions of prosody, i.e. duration and intonation, using different techniques and their subsequent integration. Our approach provides us with a flexible and extendible model that can also be used to implement, study and explain the theory behind aspects of the phenomena observed in speech prosody.

A computational model of intonation for yorùbá text-to-speech synthesis:design and analysis

In this paper we present the design and analysis of an intonation model for text-to-speech (TTS) synthesis applications using a combination of Relational Tree (RT) and Fuzzy Logic (FL) technologies. The model is demonstrated using the Standard Yorùbá (SY) language. In the proposed intonation model, phonological information extracted from text is converted into an RT. RT is a sophisticated data structure that represents the peaks and valleys as well as the spatial structure of a waveform symbolically in the form of trees. An initial approximation to the RT, called Skeletal Tree (ST), is first generated algorithmically. The exact numerical values of the peaks and valleys on the ST is then computed using FL. Quantitative analysis of the result gives RMSE of 0.56 and 0.71 for peak and valley respectively. Mean Opinion Scores (MOS) of 9.5 and 6.8, on a scale of 1 - -10, was obtained for intelligibility and naturalness respectively.

Computational model of bladder tissue based on its measured optical properties

Urinary bladder diseases are a common problem throughout the world and often difficult to accurately diagnose. Furthermore, they pose a heavy financial burden on health services. Urinary bladder tissue from male pigs was spectrophotometrically measured and the resulting data used to calculate the absorption, transmission, and reflectance parameters, along with the derived coefficients of scattering and absorption. These were employed to create a "generic" computational bladder model based on optical properties, simulating the propagation of photons through the tissue at different wavelengths. Using the Monte-Carlo method and fluorescence spectra of UV and blue excited wavelength, diagnostically important biomarkers were modeled. Additionally, the multifunctional noninvasive diagnostics system "LAKK-M" was used to gather fluorescence data to further provide essential comparisons. The ultimate goal of the study was to successfully simulate the effects of varying excited radiation wavelengths on bladder tissue to determine the effectiveness of photonics diagnostic devices. With increased accuracy, this model could be used to reliably aid in differentiating healthy and pathological tissues within the bladder and potentially other hollow organs.

A computational model of hydrogen production by steam reforming of dimethyl ether in a large scale CFB reactor. Part II:parametric analysis

This study presents a computational parametric analysis of DME steam reforming in a large scale Circulating Fluidized Bed (CFB) reactor. The Computational Fluid Dynamic (CFD) model used, which is based on Eulerian-Eulerian dispersed flow, has been developed and validated in Part I of this study [1]. The effect of the reactor inlet configuration, gas residence time, inlet temperature and steam to DME ratio on the overall reactor performance and products have all been investigated. The results have shown that the use of double sided solid feeding system remarkable improvement in the flow uniformity, but with limited effect on the reactions and products. The temperature has been found to play a dominant role in increasing the DME conversion and the hydrogen yield. According to the parametric analysis, it is recommended to run the CFB reactor at around 300 °C inlet temperature, 5.5 steam to DME molar ratio, 4 s gas residence time and 37,104 ml gcat -1 h-1 space velocity. At these conditions, the DME conversion and hydrogen molar concentration in the product gas were both found to be around 80%.

Hydrogen production by steam reforming of DME in a large scale CFB reactor. Part I:computational model and predictions

This study presents a computational fluid dynamic (CFD) study of Dimethyl Ether steam reforming (DME-SR) in a large scale Circulating Fluidized Bed (CFB) reactor. The CFD model is based on Eulerian-Eulerian dispersed flow and solved using commercial software (ANSYS FLUENT). The DME-SR reactions scheme and kinetics in the presence of a bifunctional catalyst of CuO/ZnO/Al2O3+ZSM-5 were incorporated in the model using in-house developed user-defined function. The model was validated by comparing the predictions with experimental data from the literature. The results revealed for the first time detailed CFB reactor hydrodynamics, gas residence time, temperature distribution and product gas composition at a selected operating condition of 300 °C and steam to DME mass ratio of 3 (molar ratio of 7.62). The spatial variation in the gas species concentrations suggests the existence of three distinct reaction zones but limited temperature variations. The DME conversion and hydrogen yield were found to be 87% and 59% respectively, resulting in a product gas consisting of 72 mol% hydrogen. In part II of this study, the model presented here will be used to optimize the reactor design and study the effect of operating conditions on the reactor performance and products.

Computational modelling of the impact of particle size to the heat transfer coefficient between biomass particles and a fluidised bed

The fluid–particle interaction and the impact of different heat transfer conditions on pyrolysis of biomass inside a 150 g/h fluidised bed reactor are modelled. Two different size biomass particles (350 µm and 550 µm in diameter) are injected into the fluidised bed. The different biomass particle sizes result in different heat transfer conditions. This is due to the fact that the 350 µm diameter particle is smaller than the sand particles of the reactor (440 µm), while the 550 µm one is larger. The bed-to-particle heat transfer for both cases is calculated according to the literature. Conductive heat transfer is assumed for the larger biomass particle (550 µm) inside the bed, while biomass–sand contacts for the smaller biomass particle (350 µm) were considered unimportant. The Eulerian approach is used to model the bubbling behaviour of the sand, which is treated as a continuum. Biomass reaction kinetics is modelled according to the literature using a two-stage, semi-global model which takes into account secondary reactions. The particle motion inside the reactor is computed using drag laws, dependent on the local volume fraction of each phase. FLUENT 6.2 has been used as the modelling framework of the simulations with the whole pyrolysis model incorporated in the form of User Defined Function (UDF).

Contrast integration over area is extensive: a three-stage model of spatial summation

Classical studies of area summation measure contrast detection thresholds as a function of grating diameter. Unfortunately, (i) this approach is compromised by retinal inhomogeneity and (ii) it potentially confounds summation of signal with summation of internal noise. The Swiss cheese stimulus of T. S. Meese and R. J. Summers (2007) and the closely related Battenberg stimulus of T. S. Meese (2010) were designed to avoid these problems by keeping target diameter constant and modulating interdigitated checks of first-order carrier contrast within the stimulus region. This approach has revealed a contrast integration process with greater potency than the classical model of spatial probability summation. Here, we used Swiss cheese stimuli to investigate the spatial limits of contrast integration over a range of carrier frequencies (1–16 c/deg) and raised plaid modulator frequencies (0.25–32 cycles/check). Subthreshold summation for interdigitated carrier pairs remained strong (~4 to 6 dB) up to 4 to 8 cycles/check. Our computational analysis of these results implied linear signal combination (following square-law transduction) over either (i) 12 carrier cycles or more or (ii) 1.27 deg or more. Our model has three stages of summation: short-range summation within linear receptive fields, medium-range integration to compute contrast energy for multiple patches of the image, and long-range pooling of the contrast integrators by probability summation. Our analysis legitimizes the inclusion of widespread integration of signal (and noise) within hierarchical image processing models. It also confirms the individual differences in the spatial extent of integration that emerge from our approach.

Effects of syllable structure in aphasic errors:implications for a new model of speech production

Current models of word production assume that words are stored as linear sequences of phonemes which are structured into syllables only at the moment of production. This is because syllable structure is always recoverable from the sequence of phonemes. In contrast, we present theoretical and empirical evidence that syllable structure is lexically represented. Storing syllable structure would have the advantage of making representations more stable and resistant to damage. On the other hand, re-syllabifications affect only a minimal part of phonological representations and occur only in some languages and depending on speech register. Evidence for these claims comes from analyses of aphasic errors which not only respect phonotactic constraints, but also avoid transformations which move the syllabic structure of the word further away from the original structure, even when equating for segmental complexity. This is true across tasks, types of errors, and, crucially, types of patients. The same syllabic effects are shown by apraxic patients and by phonological patients who have more central difficulties in retrieving phonological representations. If syllable structure was only computed after phoneme retrieval, it would have no way to influence the errors of phonological patients. Our results have implications for psycholinguistic and computational models of language as well as for clinical and educational practices.

Travelling-wave model of semiconductor optical amplifier based non-linear loop mirror

A travelling-wave model of a semiconductor optical amplifier based non-linear loop mirror is developed to investigate the importance of travelling-wave effects and gain/phase dynamics in predicting device behaviour. A constant effective carrier recovery lifetime approximation is found to be reasonably accurate (±10%) within a wide range of control pulse energies. Based on this approximation, a heuristic model is developed for maximum computational efficiency. The models are applied to a particular configuration involving feedback.

A compartmental CFD-PBM model of high shear wet granulation

The conventional, geometrically lumped description of the physical processes inside a high shear granulator is not reliable for process design and scale-up. In this study, a compartmental Population Balance Model (PBM) with spatial dependence is developed and validated in two lab-scale high shear granulation processes using a 1.9L MiPro granulator and 4L DIOSNA granulator. The compartmental structure is built using a heuristic approach based on computational fluid dynamics (CFD) analysis, which includes the overall flow pattern, velocity and solids concentration. The constant volume Monte Carlo approach is implemented to solve the multi-compartment population balance equations. Different spatial dependent mechanisms are included in the compartmental PBM to describe granule growth. It is concluded that for both cases (low and high liquid content), the adjustment of parameters (e.g. layering, coalescence and breakage rate) can provide a quantitative prediction of the granulation process.