Although a great deal of research on speaker recognition has been carried out over the past decades, the phonetic description of speech segments showing varying degrees of speaker specific information has not received much attention as yet. Questions that arise here are: which phonemes contain more speaker specific information than others? Is this information located in the stationary or in the transitional parts of the phoneme? Can these findings be related to coarticulation phenomena? A research project is currently under way at the University of Nijmegen which addresses these questions for a limited set of Dutch phonemes. This paper describes the analysis framework that has been developed for this project.
Cite as: Heuvel, H.v.d., Cranen, B., Rietveld, A.C.M. (1990) Inter- and intra-speaker variability in Dutch speech segments: towards an analysis framework. Proc. ESCA Workshop on Speaker Characterization in Speech Technology, 198-203
@inproceedings{heuvel90_scst,
author={Henk van den Heuvel and Bert Cranen and A. C. M. Rietveld},
title={{Inter- and intra-speaker variability in Dutch speech segments: towards an analysis framework}},
year=1990,
booktitle={Proc. ESCA Workshop on Speaker Characterization in Speech Technology},
pages={198--203}
}