Re: [simon-support] Spanish model

[Peter G. <gr...@si...>]

Am Samstag 18 Dezember 2010, 14:28:08 schrieben Sie:
> yes ! it works...
Great!

> > Building large user independant acoustic models from scratch is quite
> > complicated and takes some time and know-how. Do you have any background
> > / previous experience?
> 
> I don't have experience in voice recognition.
Then I would prepare for a bit of an adventure :)

If you don't have a specific reason, I'd recommend using the 16kHz samples (not 
the 8kHz samples) as you'll get much better results.

To create your models you can use sam (part of simon) but you'll still have to 
know a bit about the process. The Tutorials at voxforge are probably a good 
read as is the HTK book (quite complicated) and of course the simon / sam 
manual.

What is your ultimate goal in this?

Best regards,
Peter