I want to achieve a similar singing scoring function to determine the similarity of two audio, but I do not know how a simple implementation, look at a lot of github projects, more mention is simhash, but I feel similar to the audio may not be very good, so here to ask for advice.
