L'implant cochléaire est une prothèse neurale implantée visant à restituer une sensation auditive chez des personnes souffrant de surdité neurosensorielle sévère à profonde. Si les performances en reconnaissance de la parole sont relativement bonnes dans le silence, elles chutent dramatiquement dans des environnements sonores complexes. L'une des principales limites de l'appareil vient du fait que chaque électrode stimule une large portion de la cochlée. Ainsi lorsque plusieurs électrodes sont activées les champs électriques produits interfèrent ce qui détériore la transmissions des informations sonores. Plusieurs modes de stimulation ont été proposés pour remédier à ce problème mais les améliorations en termes de reconnaissance de la parole restent limités. Dans ce projet, nous cherchons tout d'abord à expliquer via une simulateur acoustique, les résultats décevants obtenus avec le mode de stimulation bipolaire. Dans un deuxième temps nous tentons de mieux comprendre le comportement électrique de l'oreille interne implantée afin d'optimiser la stimulation multipolaire phased array (van den Honert et Kelsall 2007). Pour obtenir une stimulation efficace il faut par ailleurs s'assurer de l'état de la population neuronale à stimuler. Dans ce projet nous essayons donc de mieux comprendre l'interface électrode-neurones et d'identifier un possible corrélat psychophysique de l'état des neurones. Enfin nous discutons la possibilité de créer une stimulation optimale focalisée directement au niveau des neurones. / The cochlear implant is a neural prosthesis designed to restore an auditory sensation to people suffering from severe to profound sensorineural deafness. While satisfying speech recognition can be achieved in silence, their performance dramatically drop in more complex environments. One main limitations of the present device is due to the fact that each electrode stimulates a wide portion of the cochlea. As a result, when several electrodes are activated, the electrical field produced by different electrodes overlap which distorts the transmission of sound information. Several alternative stimulation modes have been proposed to overcome this issue but the benefit in terms of speech recognition remained limited. In this project, we first used an acoustic simulator of the cochlear implant to explain the desappointing results obtained with the bipolar stimuilation mode. We then try to better understand the electrical behavior of the implanted cochlea in order to optimize the multipolar phased array stimulation strategy ( van den Honert and Kelsall 2007). To achieve an efficient stimulation of the neural population it is necessary to determine the distribution of neural survival. In this project we aim to better understand the electrode-neuron interface and identify a possible psychophysical correlate of neural survival. Finally, we discuss the main results and the possibility to design an optimal stimulation strategy to achieve a spatially-focussed electrical field at the level of the nerve fibers.

Thesis (MScIng)--University of Stellenbosch, 2004. / ENGLISH ABSTRACT: A singing voice can be manipulated artificially by means of a digital computer for the purposes of creating new melodies or to correct existing ones. When the fundamental frequency of an audio signal that represents a human voice is changed by simple algorithms, the formants of the voice tend to move to new frequency locations, making it sound unnatural. The main purpose is to design a technique by which the pitch and formants of a singing voice can be controlled independently. / AFRIKAANSE OPSOMMING: Onafhanklike formant- en toonhoogte beheer toegepas op ’n sangstem: ’n Sangstem kan deur ’n digitale rekenaar gemanipuleer word om nuwe melodie¨e te skep, of om bestaandes te verbeter. Wanneer die fundamentele frekwensie van ’n klanksein (wat ’n menslike stem voorstel) deur ’n eenvoudige algoritme verander word, skuif die oorspronklike formante na nuwe frekwensie gebiede. Dit veroorsaak dat die resultaat onnatuurlik klink. Die hoof oogmerk is om ’n tegniek te ontwerp wat die toonhoogte en die formante van ’n sangstem apart kan beheer.

<p>To effectively demonstrate the strength of using digital signal processing when producing sound effects, a sound effects demo is used at the lectures of the course TSRT78, Digital signal processing, which is given at the university in Linköping.</p><p>The amount of effects, that in an instructive way can be used for an educational purpose, are many and the existing version of the sound effects demo is somewhat limited in its range of effects.</p><p>This reports main focus lies in the presentation of what kind of effects which can be interesting in this kind of demo. All of the effects are presented with their background theory and examples on how they can be implemented in software, mainly with the focus on MATLABTM. Investigations on how well the effects can be run in realtime, in the toolbox SimulinkTM, has been made.</p><p>In the report there is also a presentation of a new version of the sound effect demo that has been produced with user friendlieness and further updates in mind. In the new demo all of the effects are implemented, according to their presentations. The report finishes with suggestions for further work on the sound effects demo.</p>

Ce travail de thèse a consisté à étudier les mécanismes perceptifs et neurofonctionnels impliqués lors de la perception de la voix chez des patients sourds postlinguaux implantés cochléaires unilatéralement, et chez des sujets normo-entendants en simulation d'implant. Pour répondre à cet objectif, nous avons testé les performances comportementales des patients implantés dans des tâches de détection de la voix ainsi que dans des tâches de perception de l'information paralinguistique de la voix, comme le genre. Les patients ont été testés au cours d'un suivi ainsi qu'en mesures transversales. Nous avons comparé leurs performances à celles de sujets normo-entendants en condition de simulation d'implant cochléaire (vocoder). Nous avons également testé les sujets normo-entendants dans un protocole IRMf consistant à mesurer l'activité spécifique à la voix lors de la simulation d'implant. Dans l'ensemble, ces travaux montrent qu'après implantation cochléaire, les patients sourds sont déficitaires en matière de perception de la voix, contrairement à la compréhension du langage. Ce déficit n'est pas uniquement lié à la dégradation du signal par le processeur de l'implant cochléaire, mais aussi certainement à des réorganisations corticales subséquentes à la surdité.

To effectively demonstrate the strength of using digital signal processing when producing sound effects, a sound effects demo is used at the lectures of the course TSRT78, Digital signal processing, which is given at the university in Linköping. The amount of effects, that in an instructive way can be used for an educational purpose, are many and the existing version of the sound effects demo is somewhat limited in its range of effects. This reports main focus lies in the presentation of what kind of effects which can be interesting in this kind of demo. All of the effects are presented with their background theory and examples on how they can be implemented in software, mainly with the focus on MATLABTM. Investigations on how well the effects can be run in realtime, in the toolbox SimulinkTM, has been made. In the report there is also a presentation of a new version of the sound effect demo that has been produced with user friendlieness and further updates in mind. In the new demo all of the effects are implemented, according to their presentations. The report finishes with suggestions for further work on the sound effects demo.

Aim : The purpose of this study was to examine the impact of simulated hearing loss on the acoustic contrasts between declarative questions and declarative statements and on the perception of speech intonation. A further purpose of the study was to investigate whether any such effects are universal or language specific. Method: Speakers included four native speakers of English and four native speakers of Mandarin and Taiwanese, with two female and two male adults in each group. Listeners included ten native English and ten native speakers of Mandarin and Taiwanese, with five female and five male adults in each group. All participants were aged between 19 and 55 years old. The speaker groups were asked to read a list of 28 phrases, with each phrase expressed as a declarative statement or a declarative question separately. These phrases were then filtered through six types of simulated hearing loss configurations, including three levels of temporal jittering for simulating a loss in neural synchrony, a high level of temporal jittering in combination with a high-pass or a low-pass filter that simulate falling and rising audiometric hearing loss configurations, and a vocoder processing procedure to simulate cochlear implant processing. A selection of acoustic measures was derived from the sentences and from some embedded vowels, including /i/, /a/, and /u/. The listener groups were asked to listen to the tokens in their native language and indicate if they heard a statement or a question. Results: The maximum fundamental frequency (F0) of the last syllable (MaxF0-last) and the maximum F0 of the remaining sentence segment (MaxF0-rest) were found to be consistently higher in declarative questions than in declarative statements. The percent jitter measure was found to worsen with simulated hearing loss as the level of temporal jittering increased. The vocoder-processed signals showed the highest percent jitter measure and the spread of spectral energy around the dominant pitch. Results from the perceptual data showed that participants in all three groups performed significantly worse with vocoder-processed tokens compared to the original tokens. Tokens with temporal jitter alone did not result in significantly worse perceptual results. Perceptual results from the Taiwanese group were significantly worse than the English group under the two filtered conditions. Mandarin listeners performed significantly worse with the neutral tone on the last syllable, and Taiwanese listeners performed significantly worse with the rising tone on the last syllable. Perception of male intonation was worse than female intonation with temporal jitter and high-pass filtering, and perception of female intonation was worse than male intonation with most temporal jittering conditions, including the temporal jitter and low-pass filtering condition. Conclusion: A rise in pitch for the whole sentence, as well as that in the final syllable, was identified as the main acoustic marker of declarative questions in all of the three languages tested. Perception of intonation was significantly reduced by vocoder processing, but not by temporal jitter alone. Under certain simulated hearing loss conditions, perception of intonation was found to be significantly affected by language, lexical tone, and speaker gender.

Thesis (M. Eng.)--Carleton University, 2000. / Includes bibliographical references (p. 94-99). Also available in electronic format on the Internet.

Thesis (M. Eng.)--Carleton University, 1991. / Also available in electronic format on the Internet.

This study focuses on improving the quality of statistical speech synthesis based on sinusoidal models. Vocoders play a crucial role during the parametrisation and reconstruction process, so we first lead an experimental comparison of a broad range of the leading vocoder types. Although our study shows that for analysis / synthesis, sinusoidal models with complex amplitudes can generate high quality of speech compared with source-filter ones, component sinusoids are correlated with each other, and the number of parameters is also high and varies in each frame, which constrains its application for statistical speech synthesis. Therefore, we first propose a perceptually based dynamic sinusoidal model (PDM) to decrease and fix the number of components typically used in the standard sinusoidal model. Then, in order to apply the proposed vocoder with an HMM-based speech synthesis system (HTS), two strategies for modelling sinusoidal parameters have been compared. In the first method (DIR parameterisation), features extracted from the fixed- and low-dimensional PDM are statistically modelled directly. In the second method (INT parameterisation), we convert both static amplitude and dynamic slope from all the harmonics of a signal, which we term the Harmonic Dynamic Model (HDM), to intermediate parameters (regularised cepstral coefficients (RDC)) for modelling. Our results show that HDM with intermediate parameters can generate comparable quality to STRAIGHT. As correlations between features in the dynamic model cannot be modelled satisfactorily by a typical HMM-based system with diagonal covariance, we have applied and tested a deep neural network (DNN) for modelling features from these two methods. To fully exploit DNN capabilities, we investigate ways to combine INT and DIR at the level of both DNN modelling and waveform generation. For DNN training, we propose to use multi-task learning to model cepstra (from INT) and log amplitudes (from DIR) as primary and secondary tasks. We conclude from our results that sinusoidal models are indeed highly suited for statistical parametric synthesis. The proposed method outperforms the state-of-the-art STRAIGHT-based equivalent when used in conjunction with DNNs. To further improve the voice quality, phase features generated from the proposed vocoder also need to be parameterised and integrated into statistical modelling. Here, an alternative statistical model referred to as the complex-valued neural network (CVNN), which treats complex coefficients as a whole, is proposed to model complex amplitude explicitly. A complex-valued back-propagation algorithm using a logarithmic minimisation criterion which includes both amplitude and phase errors is used as a learning rule. Three parameterisation methods are studied for mapping text to acoustic features: RDC / real-valued log amplitude, complex-valued amplitude with minimum phase and complex-valued amplitude with mixed phase. Our results show the potential of using CVNNs for modelling both real and complex-valued acoustic features. Overall, this thesis has established competitive alternative vocoders for speech parametrisation and reconstruction. The utilisation of proposed vocoders on various acoustic models (HMM / DNN / CVNN) clearly demonstrates that it is compelling to apply them for the parametric statistical speech synthesis.

This diploma thesis deals with pitch shifting methods of acoustical signals. The theoretic part of this thesis involves description of three different pitch shifting techniques, these are the method using a modulated delay line, PICOLA method and method using a phase vocoder. The first two methods represent the processing in time domain, the third method represents the processing in frequency domain. In relation with the PICOLA method, the thesis also mentions algorithms for pitch estimation. The practical part demonstrates the use of these methods. There is described a sampler virtual musical instrument based on the playback of the sounds stored in memory. In this part the particular units providing the required functionality are described. The generating of sounds is controlled by the MIDI protocol. In the sampler is implemented the PICOLA method.