We as a group are new to signal processing and speech recognition. We have stored .wav sounds of three different words and we are trying to record a new .wav sound and have the system compare the sound to these three pre-recorded sounds to see which one correlates the most to the newly generated sound.
We have altered the .wav sounds in various ways trying to figure out how we should tackle this problem. Thus far, we have been stuck on this issue for a while now and are only able to achieve sound comparisons with just two different sounds.
How should we tackle this problem? We would be very grateful for some general directions, thanks in advance.