Automatic Pruning
of Unit Selection Speech Databases for Synthesis without loss of Naturalness
In the paper we present our
experiments with automatic pruning of speech databases created by us for
Unit Selection based speech synthesis systems. Several algorithms have been
attempted and perceptually evaluated. An optimal size of speech database has
been reached where lose of naturalness due to unit pruning is not
perceptible.
A Genetic Algorithm for Unit
Selection based Speech Synthesis
We describe the use of a
Genetic Algorithm (GA) for the Unit Selection problem, which is essentially
a search/optimization problem. The various operators for the GA have been
defined and comparison with optimization reached by hill climbing approaches
is presented.
Unit
Selection Voice for Amharic using Festvox
In this paper, we try to
describe the issues to be considered in developing a concatenative speech
synthesizer for Amharic language. The complexity of the syllable structure
of the language, the phonetic nature of the language and the result of the
perceptual test of the synthesizer will be discussed. Comments and
recommendations for further research are included.
Experiments with Unit
Selection Speech Databases for Indian Languages
This paper presents a brief
overview of unit selection speech synthesis and discusses the issues
relevant to the development of voices for Indian Languages. We discuss a few
perceptual experiments conducted on Hindi and Telugu voices
Building Non - Native
Pronunciation Lexicon for English using a Rule based Approach
Lexicon building is a time consuming and
primarily manual process. Pronunciation Lexicons change depending upon the
accent being used in different regions. We observe that the changes in
pronunciations follow certain rules and the rules can be exploited to reuse
the existing pronunciation lexicons to build non – native pronunciation
lexicon. A rule based approach to build non – native pronunciation lexicon
for English has been proposed through this paper. Also an example based
approach to build up the rules is described.
Samachaar Vaani: A
Framework for providing Automated Spoken News Service
Several web
based news services provide news in spoken format also. The current
processes involved in recording the spoken news are time consuming and
costly. This has restricted several service providers from providing news in
spoken form. A new framework that uses automatic speech synthesis to enable
existing web based news services to provide spoken news services to the
users at no recurrent costs has been presented in this paper.
A Data - Driven
Synthesis Approach for Indian Languages using Syllable as Basic Unit
The paper describes a data driven synthesis
method for Indian Languages using Syllables as basic units for
concatenation. Unit Selection algorithm for this methods exploits the
advantages of prosodic matching function, which is capable of implicitly
selecting a larger sequence such as words, phrases and even sentences. We
show that the proposed method generates high quality speech for Indian
Languages.
Implementing a Natural
Language Conversational Interface for Indian Language Computing
Indian Language Computing is lagging behind
due to several reasons, primary amongst those being unavailability of
suitable interfaces for human – computer interaction. Natural Language
Conversational Interfaces can play the all-important role in improving the
usability of interfaces for the Indian masses. In this paper a Natural
Language Conversational Interface incorporating speech and dialog is
presented.
|