For our final project in CMU’s 36-750 Statistical Computing, I elected to build an audio fingerprinting software.
It’s currently only runnable from the command line (or a Python IDE). Ideally it would be available as a Flask app, with an audio recording function through a webpage. I may roll out an update to enable this at some point in the future, but it probably won’t come until after I’ve finished my degree.
Download, play around with, and match your audio using the GitHub repo.
This was the first of a few projects where I went after music as data rather than as listening — see also the Pitchfork review corpus.