News for hackers. Stuff that matters.
Re:Creating Shazam in Java (2010) by noipv4 on Tuesday August 23, 2022 @10:40PM (#32574368)
I thought they use Cepstrum followed by machine learning.
Re:Creating Shazam in Java (2010) by a-dub on Wednesday August 24, 2022 @01:03AM (#32575190)
windows fingerprinted by distances and angles between spectral peaks. robust to noise but not warping in time or frequency domains.

fingerprints stored in a scalable lsh style lookup table. (iirc) with the goal of a fast, scalable lookup table for all the 30s windows in all recorded music.

Re:Creating Shazam in Java (2010) by memling on Wednesday August 24, 2022 @08:13AM (#32578321)
> robust to noise but not warping in time or frequency domains.

Could you use something like a dynamic time warping algorithm for this? (I'm not super acquainted with the technique and not sure if you could get away with it in the frequency domain used for the matching.)