The notebooks here convey the concepts and the basics of Audio fingerprinting that is used in audio-matching apps like Shazam.
- Start with the notebook called "How_Audiofingerprinting_works_basic" which contains the steps for fingerprinting an audio signal.
- Then go through the notebook called "Compare_2_songs_basic" which allows you to audio fingerprint two audio signals and compare them by calculating the Jaccard similarity between their fingerprints.
- Go through the notebook called "Rank_by_similarity" to pick a track and then compare all other tracks to it and rank them by the similarity index.
- Finally, the notebook called "Compare_with_noise" allows you to pick a track, add noise to its audio signal by specifyingg it in dB, and then compare the two.
Note that you will need to have a folder named "songs" containing all the audio tracks for the notebooks to work.