the first step is to acquir the sound. u can do it by connecting the micro to your sound card and read samples of the sound continously.
the second step which is mor complecated is to identify the desired sound. for this step u must use some of digital signal processing algorithms, for example u can use the algorithmes of speech recognition