next up previous contents
Next: Phone to phone transition Up: Task 4.1: Technical Description Previous: Task 4.1: Technical Description

The model

The HMMs that we used are left to right, 3 states models, as depicted in figure 4.1. Context independent states are only used to model phone to phone transitions. We still kept context independent phone models for modelling the less co-articulated middle section of the phonetic units.


  
Figure 4.1: Phoneme model with 1 state q2 and 2 transition states q1 and q3, which are dependent of the context.
\begin{figure}
\epsfxsize=2in
\centerline{\epsffile{hmm3.eps}}
\end{figure}

To obtain a trainable system, we reduced the number of parameters to a limited number of generalized transitions by tying the transition distributions. These generalized transition models, unlike classical generalized diphone models, were not necessarily attached to a particular phone.



Christophe Ris
1998-11-10