Experiments on The Use of Demisyllables For Automatic Speech Recognition

Authors

  • G. Ruske

Abstract

The paper describes methods for an explicit segmentation of the speech signal into demisyllable segments by evaluating the output of a loudness model. Syllable nuclei are indicated by maxima of a smoothed loudness function. Consonant clusters and vowels are introduced as decision units in order to reduce the inventory of classes. Two methods for classification of consonant clusters are compared: template matching and a feature extraction approach based on acoustic cues. Sentence recognition operates on phonetic word models adapted to the demisyllable structure.

Downloads

Published

2022-12-03

How to Cite

1.
Ruske G. Experiments on The Use of Demisyllables For Automatic Speech Recognition. Canadian Acoustics [Internet]. 2022 Dec. 3 [cited 2023 Feb. 8];14(3 bis):49-50. Available from: https://jcaa.caa-aca.ca/index.php/jcaa/article/view/3517

Issue

Section

Proceedings of the Acoustics Week in Canada