Forums > MaxMSP

aka.listen/speech recognition

November 28, 2009 | 9:01 pm

Hi guys and gals,

Does anyone use speech recognition much? Namely, the aka object listen?

I am currently working with the listen object but it doesn’t recognise any words at all, apart from the odd number here and there, i.e. eight, two, etc, which were originally ‘recognised’ from either a random word or noise in the room, not the actually number being said!

This isn’t very good for what I need it for. Has anyone else used this object and had the same problems, or know the way to correct/improve it?

Are there any other speech recognition tools I can use that are more efficient than aka.listen?

Thanks in advance :)


November 28, 2009 | 9:29 pm

You can use the built in voice recognition on a mac with some applescripts to trigger key presses which you can then intercept in max and from there you can do whatever you want. It is a kind of roundabout method but it might be useful. I explained it here in a previous post. I hope the link works.

lh


Viewing 2 posts - 1 through 2 (of 2 total)