In limited vocabulary speaking systems where several pre-recorded speech items are concatenated to generate the message one of the main problems is to add new items to the message set recorded formerly. This procedure is complicated, expensive and the new message(s) cannot be integrated into the original system without leaving the impression that the new items were generated during a later process. In most cases the new items have different tempo, voice timbre and fundamental frequency. Application users would like to have tools which eliminate these problems. To meet these requirements a method and a tool (for Hungarian) has been developed which gives the possibility to create limited vocabulary but open systems. Using this method new speech items can be generated (without recording) with the same voice as that of the original system, so the user will not feel that new message items are combined with the originally recorded items.
Cite as: Olaszy, G., Németh, G., Olaszi, P., Gordos, G. (1999) Interactive, TTS supported speech message composer for large, limited vocabulary, but open information systems. Proc. 6th European Conference on Speech Communication and Technology (Eurospeech 1999), 943-946, doi: 10.21437/Eurospeech.1999-230
@inproceedings{olaszy99_eurospeech,
author={Gábor Olaszy and Géza Németh and Péter Olaszi and Géza Gordos},
title={{Interactive, TTS supported speech message composer for large, limited vocabulary, but open information systems}},
year=1999,
booktitle={Proc. 6th European Conference on Speech Communication and Technology (Eurospeech 1999)},
pages={943--946},
doi={10.21437/Eurospeech.1999-230}
}