VoxForge
VoxForge
Hi ken, Offir,
I'd like to help but unfortunately I can't find the way . . .
I'm definitely sure, generating a new Hebrew submissions / Acoustic Models wouldn't be a wise because as I saw, too much time invested at current Hebrew submissions / Acoustic Model.
I'd like to further the Hebrew submissions/Acoustic Model!
Thanks,
Eyal
Hi Eyal,
> Can I help?
yes!
>Is someone today managing the Hebrew submissions/Acoustic Model?
Offir provided translations for the speech submission app. So that deals with the collection of speech. To create an Acoustic Model, the next step is the creation of a pronunciation dictionary (see Step 2 of the VoxForge Tutorial; one of my posts on this thread provides additional information).
If you are interested in helping out, a pronunciation dictionary would be a good task to complete.
>What is the current Hebrew Acoustic Model status (2008-03-29)?
There is no Hebrew acoustic model.
We are only at the very beginning. Translations for the Java speech submission application is the first step. Next is the creation of a pronunciation dictionary.
> Is it possible to implement the Hebrew Acoustic Model?
Yes, but you need a few hours of speech and a pronunciation dictionary to create a monophone acoustic model.
Ken