This paper proposes a speech synthesis technique integrating training and synthesis processes based on the Bayesian framework. In the Bayesian speech synthesis, all processes are derived from one single predictive distribution which represents the problem of speech synthesis directly. However, it typically assumes that the posterior distribution of model parameters is independent of synthesis d...

As speech synthesis techniques become more advanced, we are able to consider building high-quality voices from data collected outside the usual highly-controlled recording studio environment. This presents new challenges that are not present in conventional text-to-speech synthesis: the available speech data are not perfectly clean, the recording conditions are not consistent, and/or the phonet...

This paper deals with a speech synthesis. Speech synthesis is an artificial production of speech by a computer. Speech synthesis from a text is called Text-to-Speech (TTS) synthesis. The main goal of this paper is to propose and implement a TTS system based on the MBROLA (Multiband Resynthesis Overlap-Add) project. We implement jSynt tool, a TTS system written in java. The first version include...

In a recent paper we proposed a possible generalization of L. Schwartz’s classical spectral synthesis result for continuous functions in several variables. The idea is based on Gelfand pairs and spherical functions while “translation invariance” is replaced by invariance with respect to the action of affine groups. In this paper we describe the function classes which play the role of the expone...

This paper discusses two well-known phonetic-phonological phenomena of Japanese, vowel devoicing and mora-timed rhythm. Regarding vowel devoicing as a process of vowels getting consonantal, it may change the mora-templated syllable structures of Japanese, resulting in the deviation from mora-timed rhythm. To investigate spontaneous Japanese with comparison to other languages, the label data of ...

Both digital waveguide and finite difference techniques are numerical methods that have been demonstrated as appropriate for acoustic modelling applications. Whilst the application of the digital waveguide mesh to vocal tract modelling has been the subject of previous work, the application of comparable finite difference techniques is as yet untested. This study explores the characteristics of ...

In recent years the application of computer software to the learning process has been acknowledged an indisputably effective tool supporting traditional teaching methods. A particular focus has been put on the application of computational techniques based on speech and language processing to second language learning. At present, a number of commercial self-study programs using speech synthesis ...

We present a novel approach to 3D shape synthesis of closed surfaces. A curved or polyhedral 3D object of genus zero is represented by a curvature distribution on a spherical mesh that has nearly uniform distribution with known connectivity among mesh nodes. This curvature distribution, i.e., the result of forward mapping from shape space to curvature space, is used as the intrinsic shape repre...

In this letter we synthesize numerically the Lü attractor starting from the generalized Lorenz and Chen systems, by switching the control parameter inside a chosen finite set of values on every successive adjacent finite time intervals. A numerical method with fixed step size for ODEs is used to integrate the underlying initial value problem. As numerically and computationally proved in this wo...

This paper describes a method of analysing words through morphological decomposition when the lexicon is incomplete. The method is used within a text-to-speech system to help generate pronunciations of unknown words. The method is achieved within a general morphological analyser system using Koskenniemi twolevel rules.