Yes, it essentially just runs a perceptual hash algorithm on the audio and finds the closest matches in a database.