Objetivo
Development of multi-voice and multi-language Text-to-Speech (TTS) and Speech-to-Text (STT) conversion system (languages: Belorussian, Polish, Russian) Field: Computer Science Sub field: Artificial Intelligence, Signal & Image processing, Pattern Recognition (0506). Free words: Speech Recognition, Speech Synthesis, Speech Communication, Speech Technology, Speech Signal Processing, Natural Language Processing.INTAS Funding: 80,000Starting date: 1 March 2005Duration: 24 months Coordinator: Prof. Dr.-Ing. habil. Rudiger Hoffmann
Objectives:
This research aims to fill the gap in introducing and promoting computerized speech technology for Slavonic languages, in particular for Belarusian, Polish and Russian.
Research activities:
Research activities in the development of Belorussian, Polish, Russian linguistic resources and multi-voice acoustical database are aimed at providing lexical, morphological and syntactic level information, including words stress position, grammatical category of words; studio digital sound recording data, including phonemic transcription and prosodic labeling of four male and female voices, and two types of reading text providing
a) voice and phonemic peculiarities,
b) prosodic features of speech.
Expected Results:
The research and development work done within the scope of the project includes the creation of reusable Belorussian, Polish and Russian linguistic resources and multi-voice acoustical databases, methodologies and algorithms of multi-lingual TTS synthesis, voice cloning technology and STT recognition as well as the construction of their pilot application prototypes.
Convocatoria de propuestas
Data not availableRégimen de financiación
Data not availableCoordinador
D-01062 Dresden
Alemania