Abstract
In this paper, we describe a method for extracting speech segment features from an audio recording such that speech may be reconstructed while maintaining emphasis and stress with minimal annotation. We also present a script for the linguistic research tool, Praat, for performing this extraction and reconstruction via KlattGrid synthesis [1]. This technique contributes to the ongoing research efforts to preserve the Betine language (ISO 639-3:eot), one of 10 critically endangered languages spoken in Côte d’Ivoire, West Africa.
Recommended Citation
LOUKUSA, HANNAH and HAAPOJA, JACOB
(2024)
"KLATTGRID AND FORMANT-BASED SPEECH SYNTHESIS: DEMONSTRATION WITH BETINE,"
Linguistic Portfolios: Vol. 13, Article 9.
Available at:
https://repository.stcloudstate.edu/stcloud_ling/vol13/iss1/9
