Speech Segmenting and kinematics

Auteurs-es

  • J Caelen

Résumé

Conclusion:

The above makes it possible to look at segmenting, and subsequently at acoustico-phonetic decoding, from a new and maybe more advantageous angle instead of researching discontinuity, we would resort to the formal instruments of mechanics (or data-analysis) to examine local variations in speech-trajectories that are represented in suitable spaces. Such a representation allows for an ascending description, from acoustics to phonology; while by-passing any a priori (even implicit) phonetic model. At the same time, it seems possible to find a grammar of distortions capable of superposing the several trajectories that correspond to one sequence uttered by several speakers. This kind of results, nevertheless, remains to be confirmed over large speech-corpuses and large numbers of speakers,

Fichiers supplémentaires

Publié-e

2022-12-03

Comment citer

1.
Caelen J. Speech Segmenting and kinematics. Canadian Acoustics [Internet]. 3 déc. 2022 [cité 21 nov. 2024];14(3 bis):77-9. Disponible à: https://jcaa.caa-aca.ca/index.php/jcaa/article/view/3550

Numéro

Rubrique

Actes du congrès de la Semaine canadienne d'acoustique