Matches audio to small vocabulary using fast fourier transforms - View it on GitHub