Component Structuring and Trajectory Modeling for Speech Recognition - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2014

Component Structuring and Trajectory Modeling for Speech Recognition

Denis Jouvet


When the speech data are produced by speakers of different age and gender, the acoustic variability of any given phonetic unit becomes large, which degrades speech recognition performance. A way to go beyond the conventional Hidden Markov Model is to explicitly include speaker class information in the modeling. Speaker classes can be obtained by unsupervised clustering of the speech utterances. This paper introduces a structuring of the Gaussian compo- nents of the GMM densities with respect to speaker classes. In a first approach, the structuring of the Gaussian components is combined with speaker class-dependent mixture weights. In a second approach, the structuring is used with mixture transition matrices, which add dependencies between Gaussian components of mixture densities (as in stranded GMMs). The different approaches are evaluated and compared in detail on the TIDIGITS task. Significant improvements are obtained using the proposed approaches based on structured components. Additional results are reported for phonetic decoding on the NEOLOGOS database, a large corpus of French telephone data.
Fichier principal
Vignette du fichier
inter2014_agorin_v11.pdf (206.82 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01063653 , version 1 (12-09-2014)


  • HAL Id : hal-01063653 , version 1


Arseniy Gorin, Denis Jouvet. Component Structuring and Trajectory Modeling for Speech Recognition. Interspeech, Sep 2014, Singapoore, Singapore. ⟨hal-01063653⟩
380 View
161 Download


Gmail Facebook Twitter LinkedIn More