I am both new to linux and "sound processing" and I'm sorry if this looks like a question being answered with the first entry on Google. Because I don't even know what to google for.
A friend of mine and I are trying to program an intelligent personal assistant that should do all kinds of stuff for us like google stuff, play music, tell us jokes... But that doesn't matter at this point.
The voice will be translated into text and text will be translated to speech via Google translate API which doesn't cause too much trouble. However our mic is kind of bad and we want to improve the quality of our sound input e.g. remove background noise and/or filter our voices from other disturbing noises. It would be brilliant if that would happen on the fly.
We are using Ubuntu and Raspbian at the moment.