Story perspectives

Cambridge Researchers Use AI to Decode Jazz Pianists’ Styles

8/18/2026

Summary

Story summary

Three University of Cambridge researchers used AI to identify the playing style of 20 celebrated jazz pianists.
They trained models on 84 hours covering 1,629 performances by Chick Corea, Thelonious Monk, Oscar Peterson and Bill Evans.
One model identified the pianist correctly in more than 94 percent of tests, while a second model achieved 91 percent accuracy.
The researchers released a web app that visualizes each pianist’s fingerprint with graphs and audio samples.

1 of 1

Story summary

Three University of Cambridge researchers used AI to identify the playing style of 20 celebrated jazz pianists.
They trained models on 84 hours covering 1,629 performances by Chick Corea, Thelonious Monk, Oscar Peterson and Bill Evans.
One model identified the pianist correctly in more than 94 percent of tests, while a second model achieved 91 percent accuracy.
The researchers released a web app that visualizes each pianist’s fingerprint with graphs and audio samples.