i would like to now if MARF is capable of some things i'd like to do. If i get some feedback here, this would be great. Thanks.
My scenario would be live identification. So first: Would MARF be able to perform on an audiostream that is directly recorded from the Mircrophone? If so, what would be a good windowsize (in seconds) to make a reliable prediction? Further, would it be possible to get the number of speakers, or to detect that the speaker has changed. And last but not least, would i be able to detect a speaker that is not in the training set as "Unknow Speaker" rather than missclassfied?
Thank you so much.
I read that:
Real-Time recording and recognition
is in todo list: http://marf.sourceforge.net/docs/marf/0.3.0.6/TODO
About "Unknow Speaker" I too have this problem. Have you solved?
I didn't solve it. Sorry. But i am still interested.
Ok, then we can cooperate for to do this. Are you agree?
Yes, I agree, we can cooperate for to do this.
I guess that wasn't me, but some jester.
And i am sorry, but i don't think i'll have time to digg into this.