Apple introduces ShazamKit framework for custom audio recognition

Apple has introduced ShazamKit, a new framework that allows developers to integrate song recognition capabilities into their own applications. While the tool utilizes Shazam’s existing extensive catalog, it also enables the creation of custom audio catalogs.

This functionality allows for the detection of specific audio signatures beyond standard music, such as particular podcast episodes or specific clips from television shows. This distinguishes ShazamKit from general machine learning tools like MLSoundClassifier, which are better suited for ‘fuzzy matches’ like voice recognition, whereas ShazamKit is designed for identifying precise, predefined audio signatures.