speech recognition

Hello all,

I've been handed a task for a third year project requiring me to do voice recognition [differentiate between people and their voices and identify them] by means of signal processing (I'm assuming ill be working heavily with frequency content). I know this might seem vague and not really to the point, because I'm fairly new to this all (android and also applied digital signal processing [only studied theory thus far]), but I was hoping I could be pointed to some API and maybe some general tips on how I might want to approach this from someone who might have experience with audio processing on android. Also from what I read android is forward compatible and that would mean I'd have to find out which version would be suited for me and my project requirements in terms of API, and that would imply that I should go with the lowest version possible that would meet my requirements since I'd like to target as many phones as possible that are running android. I'd greatly appreciate any and all suggestions and recommendations. Best regards